A review of the FBI's Handling:Chapter 5 IV

IV. OIG’s analysis of the FBI’s handling of the intelligence information concerning Hazmi and Mihdhar
We found systemic and individual failings in the FBI’s handling of the Hazmi and Mihdhar matter. As a result of these failings, there were at least five opportunities for the FBI to connect information that could have led to an earlier investigation of Hazmi and Mihdhar and their activities in the United States. In this analysis section, we first discuss the systemic problems involving the breakdowns in the gathering or passing of information about Hazmi and Mihdhar between the FBI and CIA. We then turn to the problems in handling intelligence information within the FBI. Finally, we discuss the actions of individual FBI employees in handling information about Hazmi and Mihdhar information. In this section, we do not make recommendations regarding the actions of the CIA and its employees. We believe the CIA shares a significant responsibility for the breakdowns in the Hazmi and Mihdhar case, and that several of its employees did not provide the intelligence information to the FBI as they should have. We leave it to the CIA OIG, the entity with oversight jurisdiction over the CIA and its employees, to reach conclusions and make recommendations on the actions of the CIA and its employees. 316 A. Systemic impediments that hindered the sharing of information between the CIA and the FBI The most critical breakdown in the Hazmi and Mihdhar case was the failure of the FBI to learn from the CIA critical information about them; their travel to the United States; and their association with Khallad, the purported mastermind of the Cole attack. These breakdowns reflected serious problems in the process before the September 11 attacks for sharing information between the FBI and the CIA. The FBI failed to receive from the CIA three critical pieces of intelligence about Mihdhar and Hazmi in a timely manner: • Mihdhar’s possession of a valid, multiple-entry U.S. visa; • Hazmi’s travel to the United States; and • The identification of Khallad in a surveillance photograph of the Malaysia meetings attended by Hazmi and Mihdhar and other al Qaeda operatives in January 2000. The CIA became aware of these three pieces of intelligence in January 2000, March 2000, and January 2001. Despite claims to the contrary, we found that none of this information was passed from the CIA to the FBI until August 2001. Although the CIA failed to timely pass this information to the FBI, there were several opportunities for the FBI to have obtained this information in other ways. But significant systemic problems, which we describe below, hindered the flow of information between the CIA and the FBI. 1. Use of detailees One of the most significant opportunities for the FBI to have obtained the intelligence information relating to Hazmi and Mihdhar was through the FBI detailees at the CTC. As discussed above, the FBI detailees to the CTC had access to CIA cable traffic and could read the cables that discussed Mihdhar’s U.S. visa, the surveillance of the meetings of al Qaeda operatives in Malaysia, Hazmi’s subsequent travel to the United States, and the Khallad identification from the Kuala Lumpur photographs. Several of the FBI detailees accessed and read some of these cables. Significantly, in January 2000, one detailee, Dwight, prepared a draft CIR to pass to the FBI the information about Mihdhar’s visa, his al Qaeda connections, and his travel to Malaysia. The FBI 317 should have been informed of this information because of its clear domestic nexus. However, the CIR was never sent to the FBI. According to a note on the CIR, John, a Deputy Chief of the Bin Laden Unit, directed that the CIR be placed on hold, and FBI detailees did not have authority to disseminate CTC information without approval from the CIA. Eight days later, Dwight inquired about the disposition of the CIR through an e-mail to John asking whether anything needed to be changed on the cable. However, this e-mail failed to prompt further action on this CIR. The witnesses we interviewed had no recollection of the CIR and why it was not sent. We found no further record that anything was done with regard to the CIR. In our view, the CIA should have sent the CIR to the FBI because of the important information it contained, and the FBI detailee should have followed up to ensure that it was sent. While we found evidence that Dwight inquired about its status at least once, there is no evidence that he took any other action to ensure that the information was sent to the FBI, including inquiring with other CTC supervisors about the need to send the cable to the FBI. In reviewing the actions of the detailees, we found that the FBI lacked clear guidance on the role and responsibilities of FBI detailees to the CTC’s Bin Laden Unit. This led to inconsistent expectations about what they were supposed to be doing at the CTC. Our review of the documents and interviews with the five FBI detailees to the CTC’s Bin Laden Unit found that none of them had defined duties that were clearly understood, either by them or FBI managers. Nor were there any memoranda of understanding (MOU) between the FBI and the CIA setting out the job duties and responsibilities of any of the detailees.257 Moreover, we asked the FBI for the performance appraisals for all five of the detailees to the Bin Laden Unit during this period, and we received 257 We asked both the FBI and the CIA for any memoranda of understanding between the agencies specifying the job duties of any of the detailees. The only MOUs we received, which were provided by the CIA, related to the administrative nature of the details, such as time and attendance reports, travel and training expenses, security clearances, and medical coverage. The MOUs did not address their substantive duties or responsibilities. 318 appraisals for three of them. They revealed that the FBI detailees were evaluated based on the elements for their positions at the FBI, not based on whatever they were supposed to be doing while working at the CTC.258 The FBI was unable to provide any other documents defining or outlining the roles or responsibilities of these detailees. We also interviewed the detailees about their understanding of their roles and responsibilities at the CTC. They stated that they were not given any specific instructions about their job duties. They described their details at the CTC as ill-defined and with little direction. As a result, each detailee defined the job at the CIA as he or she determined it to be, and there was significant variation in their conceptions of the job. For example, Dwight told the OIG that he focused on leads that were related to financial components of terrorism, which he developed from various sources, such as from reviewing cable traffic, from his supervisors at the CTC, and from referrals from CIA officers at the CTC. By contrast, Malcolm told the OIG that he thought he was the “eyes and ears” of the New York Field Office, and that his role was “to monitor” cases being worked jointly by the CIA and the New York Field Office, such as the East African embassy bombings investigation. He said that he also would follow up on requests for information from the FBI to the CIA. Moreover, Mary said she was not given any specific instructions about her role at the CIA, but she was eventually trained to be a CTC desk officer and that was how she operated – like other CTC desk officers with specific assignments or “accounts.” Eric, who was a Bin Laden Unit Deputy Chief, said that he was told “to fix” the relationship between the Bin Laden Unit and the FBI, but he was not given any specific instructions about how to go about accomplishing this objective. He said that he assisted in the running of the Bin Laden Unit by directly overseeing CTC operations and that he also functioned in a liaison role between the CIA and the FBI. He supervised the FBI detailees like he did other Bin Laden Unit employees. He was not given any other supervisory 258 For a fourth detailee, Mary, the FBI produced only a performance plan but no appraisal reports. The performance plan was related to her duties as an FBI IOS. Mary told the OIG that she was directed by CTC management based on her work as a CIA desk officer and was not evaluated by FBI personnel. 319 oversight particular to the detailees. He said that on his own initiative he tried to stay abreast of matters that might be of interest to the FBI by reading the CTC cable traffic. However, he explained that determining what might be of interest to the FBI was very subjective because there were no criteria defining what should be brought to the attention of the FBI. We also interviewed the highest-ranking FBI employee detailed to the CTC, who was a Deputy Chief of the CTC from 1999 through 2002. We call him “Evan.” Evan believed that one of the FBI detailees’ functions would have been to review CIA cable traffic for information of potential relevance to the FBI. Yet, the detailees told the OIG that while reviewing CIA cable traffic was part of their jobs, it was not their function to review cable traffic for items of interest to the FBI, and they did not review all of the cable traffic on a daily basis. They said they did not think they were acting as backstops to ensure that anything that might be relevant to the FBI was brought to the FBI’s attention.259 The detailees asserted emphatically that their function did not entail scouring CIA cable traffic for the FBI, and their efficacy would be limited if they were perceived by CIA personnel merely as moles for the FBI.260 They also explained that even if this had been their role, it would have been difficult to do because of the volume of cables, especially during the chaotic Millennium period. The two FBI employees who held similar supervisory positions – one as a deputy chief in the Bin Laden Unit and the other as a deputy chief in another unit that later housed the Bin Laden Unit – also had differing views on their responsibility for reviewing cable traffic. Both agreed that their role was not merely to review cable traffic for items of interest to the FBI. Eric told the 259 We also interviewed the first FBI employee detailed in March 1996 to Bin Laden Unit soon after it was created. This detailee was an agent from the FBI’s New York Field Office, and he remained at the CTC until August 1998. He said that he did not attempt to review all of the cable traffic. He indicated, however, that when he did locate information of interest to the FBI, he did not encounter problems obtaining the CIA’s permission to share this information with the FBI. 260 Some CIA employees we interviewed stated that they, by contrast, believed that this was the function of the New York Field Office detailee. We discuss this further in the next section. 320 OIG that while he tried to review the traffic in order to stay abreast of the information in the CTC, it was too much for one person to manage effectively. By contrast, Craig, who followed Eric as a manager detailed to the CTC, told the OIG that he did not even attempt to review the cable traffic but only focused on those cables that required action on his part. In addition to failing to clearly define the roles and responsibilities of the detailees, the FBI did not provide oversight of the detailees. Eric acted as one of two deputy chiefs within the Bin Laden Unit. After Eric left the CTC, Craig was a deputy chief in a much larger unit that included the Bin Laden Unit. Both said that they performed day-to-day supervision of the detailees in the same manner in which they supervised the other CTC employees assigned to their groups.261 According to Eric and Craig, they did not focus specifically on the role of FBI detailees. Evan told the OIG that he did not supervise any of the detailees, and he had no authority to oversee their duties or direct their activities, except by virtue of his position as a senior manager within the FBI. He said that they were evaluated by their chain of command in the FBI office from which they had been assigned, which is supported by the limited documents we reviewed. We found that there was no oversight by the FBI of the detailees based on their function as detailees. The FBI’s failure to adequately oversee the detailees is illustrated by the role of Mary, the only FBI analyst detailed to the Bin Laden Unit. She has been detailed to the CIA since 1998. Mary had the opportunity to learn valuable analyst skills by working alongside CTC personnel and then use those skills at the FBI. Additionally, the detail provided an opportunity to learn about the CIA infrastructure and establish liaison contacts at the CIA. Mary told us that she operated as a full-fledged CIA desk officer, and that she has worked with FBI personnel during her detail but from the position of a CIA employee, not an FBI employee. We believe there needs to be a review of the duration of these details to ensure the value of these details is maximized. 261 Eric left the CTC in mid-January 2000, and Craig did not arrive at the CTC until July 2000. Thus, between mid-January and July 2000 the FBI had no supervisory presence for the FBI employees detailed to work Bin Laden matters at the CTC. 321 At a time when the FBI is concerned about the shortage of qualified analysts to do the work it has, a 5-year detail of an FBI analyst working as a CTC employee warrants review by the FBI.262 The same lack of oversight and direction was evident regarding the work of Malcolm, the FBI New York Field office detailee to the CTC. He had been traveling to the CTC from New York on a weekly basis for four years, until January 2003. On Mondays he traveled from New York to the CTC, stopping by FBI Headquarters. On Fridays he stopped by FBI Headquarters on his way back to New York. After the bombing of the Cole, he spent at least half of his days in Washington, D.C. at FBI Headquarters. Thus, he was frequently away from the CTC and not in a position to maximize his potential for obtaining information at the CTC. This also left the perception with other CTC employees that he was not fully integrated into the CTC. We found that that the FBI lacked a systematic approach to its use of detailees at CTC’s Bin Laden Unit. The detailees could have functioned in one of three ways – as fully integrated members of the CTC working unilaterally on CTC matters, as backstops ensuring all pertinent CTC information was forwarded to the FBI, or in some combination thereof. While there are potential benefits to using the detailees in any of these functions, the potential benefits were not maximized because there was no clear understanding of the detailees’ roles and no system to ensure that any objectives were met. The lack of oversight over FBI detailees to the CTC resulted in squandering critical opportunities for information sharing between the CIA and FBI. We also found significant misunderstandings between employees of these two agencies regarding their respective responsibilities for information sharing. First, as noted above, we found that some CIA employees believed that FBI detailees had more responsibility for reviewing the CIA cable traffic than the FBI detailees believed that they had. One CIA Bin Laden Unit employee told the OIG that the CIA was not going to “spoon feed” information to the FBI and that the FBI personnel at the Bin Laden Unit had access to all of the CIA cable traffic. She stated that while the CTC provided to the FBI intelligence 262 The OIG is in the process of completing a comprehensive review of FBI’s analyst program. 322 information that contained a domestic nexus, she did not believe it was the CIA’s responsibility to provide all of the predicating material, since the FBI detailees also had access to the same cables. In addition, CIA personnel described FBI detailee Malcolm as a “mole” for the FBI’s New York Office, suggesting they thought he was reading CIA cables for the express purpose of reporting back to the New York Field Office on what he found. In addition, we found that a similar misunderstanding existed among FBI employees in New York with respect to the role of the CIA employee detailed to the FBI’s New York Field Office. A CIA employee assigned to the JTTF in the New York Field Office had a desk in that office’s sensitive compartmented information facility (SCIF).263 FBI agents in the New York Field Office asserted to the OIG that this individual was knowledgeable regarding their investigations and that he was responsible for reviewing CIA traffic, finding items of interest to the FBI, and bringing this information to the attention of appropriate New York agents. The CIA employee, however, denied that this was his role. He told the OIG that he had been sent to the New York Office to “improve the relationship between the CIA and the FBI” and that he provided the FBI with CIA intelligence that was designated for the FBI New York Field Office’s review. He stated, however, his job was not to “spoon feed” information but only to make it accessible to the agents in New York. This meant that he would print information obtained from CIA databases that was of potential interest to the FBI New York Field Office and make that information available for review in the SCIF if FBI agents decided to come and review it. But, apparently unknown to many New York FBI agents, he believed the onus was on FBI personnel to come into the SCIF and see if any new, relevant information had arrived, rather than to alert them to that information. He also said that while he generally knows what the various FBI squads are investigating, the New York JTTF has over 300 members and he could not reasonably be expected to have knowledge of all their investigative interests. He said that if he spent his time 263 The FBI agents do not routinely work in a SCIF area. The computers on which they access ACS do not contain sensitive compartmented information or materials classified above Secret. Because a high percentage of CIA traffic contains this information, the CIA detailee must work in a separate area. 323 solely looking for information of interest to the FBI, he would never get any work done. As a result, FBI agents in New York believed they were receiving from this CIA employee assigned to the JTTF all of the CIA information of interest to the FBI, when in fact they were not. Therefore, the New York agents could have received information on Hazmi and Mihdhar directly through their own CIA employee, but they misunderstood the process. 2. FBI employees’ lack of understanding of CIA reporting process These gaps in the information sharing process were exacerbated by FBI personnel’s lack of understanding of the CIA’s reporting process. This problem is clearly illustrated by the failure of the FBI to obtain the information on the identification in January 2001 of Khallad in the Kuala Lumpur photographs by the joint FBI/CIA joint source. As detailed above, we concluded that the FBI’s ALAT was not made aware of the source’s identification of Khallad in the Malaysia meetings photographs. Although the ALAT attended the debriefing of the source, the ALAT did not immediately receive the information that the source had identified Khallad. We were unable to ascertain the reasons for this significant omission. However, our review found that there were later opportunities for the ALAT to have obtained information about the identification from CIA documents. In addition, we found that the New York FBI agents working the Cole attack investigation did not learn of this significant information, despite interviewing the source on several occasions. We believe this was due in part to the fact that the FBI personnel were not familiar with the CIA’s process for reporting intelligence information. As discussed previously, the CIA primarily relies on cable traffic to share intelligence among its personnel who are stationed around the world. None of these cables are available for FBI review, except by the limited number of FBI personnel with direct access to CIA computer systems, such as the detailees at the CTC. The CIA uses a certain type of cable called a TD to disseminate CIA information outside of the CIA to other U.S. government agencies. These cables are created by CIA reports officers based on their review of the internal 324 CIA cable traffic. The reports officers were described to us as “editors” who remove references to sources and methods contained in the cables and determine what information should be further disseminated in the TDs. As a result, TDs did not necessarily include all the substantive information contained in the internal cable traffic. Our review found the ALAT did not understand that the TDs did not necessarily contain all of the intelligence gathered by the CIA from a particular source or on a particular event. The ALAT had been keenly aware of the significance of Khallad to the FBI, and contemporaneous FBI documents outline his efforts in mid-January 2001 to try to ensure that all the information obtained from the joint source was provided to the UBL Unit at FBI Headquarters and the Bin Laden Squad in the New York Field Office. However, he relied on the TDs concerning the source’s reporting to ensure the completeness of the information that he had provided to his FBI colleagues. The ALAT erroneously believed he had obtained all the source reporting through the TDs. This was not the case. The January 2001 Khallad identification was only reported in an internal CIA cable and was never included in a TD. In addition to the ALAT, New York FBI agents working on the Cole investigation told us that when they read a TD regarding a particular subject (which they could access through CTLink), they mistakenly believed that it contained all relevant information from the source debriefings. The primary Cole case agent told us that he believed that the CIA operational cables dealt with techniques and methods, but he did not know that these cables also contained the details of debriefings. He said that he had “assumed” all the substantive reporting would be contained in the TDs, so he never asked the CIA to allow him to review the underlying cable traffic. If these FBI employees had a more thorough knowledge of the information flow within the CIA, they could have ensured that they received all the relevant information from the joint source. This was especially significant in the case of Hazmi and Mihdhar because the CIA and FBI had decided the majority of the joint source’s reporting would be handled through CIA channels, and the ALAT did not independently report in FBI documents most of the source’s information. For example, in this case, the FBI could have requested to review the CIA’s internal cables or asked the interviewing CIA officer to review the TDs and the FBI documentation to ensure all the 325 information had been captured. However, the lack of understanding by FBI personnel of the CIA reporting process and its procedures for sharing intelligence contributed to the FBI not learning of significant information in CIA cables about Khallad – which would have tied an al Qaeda operative to the Malaysia meetings attended by Mihdhar and potentially resulted in the FBI focusing on Mihdhar much earlier. 3. Inadequate procedures for documenting receipt of CIA information We also found that the FBI lacked consistent policies or procedures for the receipt and documentation of intelligence information received from the CIA. In addition, structural impediments within the FBI undermined the appropriate documentation of information received from the CIA. As we detailed above, the information concerning the surveillance of suspected al Qaeda operatives at the Malaysia meetings, including Mihdhar, was verbally conveyed in January 2000 by a CIA officer to two FBI employees who were working in the FBI’s Strategic Information Operations Center (SIOC). But this important information was not documented in any retrievable form at the FBI. The FBI was able to provide only three documents regarding the briefing on this information. First, one FBI e-mail message was recovered through a painstaking review of messages on an FBI server that the FBI searched in connection with a request from the JICI. Although this written record survived from that time, no analyst or agent would have had access to the information, learned of its existence, or been able to conduct the type of search that led to the discovery of this document. Second, information regarding the briefing was also located in one of the FBI Director’s daily briefing documents prepared in response to the Millennium threats. These briefing documents, however, were not electronically archived in a searchable database that analysts or agents in the field could access. Third, a brief handwritten note about the information he received from the CIA was contained in the personal daily calendar of one of the FBI employees briefed by the CIA officer in the SIOC. We found there were no clear procedures for documenting intelligence communicated by the CIA to the FBI in an informal manner, such as the verbal 326 briefings on Mihdhar in the SIOC. Although the SIOC had been activated during the Millennium for the express purpose of handling threat information from various sources, FBI personnel assigned to the SIOC during this period told us that there were no procedures for the receipt and handling of interagency information communicated informally unless it related to an ongoing FBI investigation. Although one witness suggested that some type of log might have existed to record incoming physical information, such as documents, the FBI found no such log. Moreover, FBI witnesses told us that the log would not have been used to document verbal briefings. Therefore, any documentation of information received informally would have been at the discretion of the recipient. We are not suggesting that every informal communication from the CIA to the FBI should be documented. We also recognize it is difficult to know the significance of any individual piece of information when it is received. Yet, we believe that the FBI should attempt to establish criteria or guidance for determining what information from informal briefings should be documented, and how it should be documented. The information received in the SIOC on Mihdhar was recorded only in a briefing provided to the Director and executive staff, which is not available to others throughout the FBI. Clearly, the authors of the Director’s daily briefing believed there was some import to this information. Because the Mihdhar information was never documented in an accessible format, only those individuals personally informed about the CIA’s information on the Malaysia meetings or those present for the Director’s briefings were made aware of the Mihdhar information. In effect, it was lost to everyone else because no analysts or field agents would be able to search for or locate this information. An effective analytical program requires that analysts have access to all available information, and that pertinent information is not contained solely in the personal memories of selected individuals. This was particularly significant because the information on Mihdhar initially did not appear to be important. But it subsequently became very significant. In the summer of 2001, FBI personnel eventually recognized the significance of the Malaysia meetings. At this time, the e-mail and the information from the Director’s briefing in January 2000 were not available to the FBI personnel. Without mechanisms to maintain information in which the significance is not immediately apparent, the FBI will not be able to fully connect and analyze disparate pieces of information for their significance. 327 In addition, even if the agents who received the information in the SIOC had wanted to document it in a form that was available throughout the FBI, the FBI lacks an information technology system capable of adequately handling this type of information. As discussed previously, the FBI’s primary electronic information storage system is the Automated Case Support (ACS) System. ACS is a case management system designed to capture information related to specific investigations and not for this type of general intelligence information. There was no FBI system that would allow this type of information to have been maintained so that it would be available for directed searches or other subsequent data mining. It is also important to note that ACS is not approved for storage of information classified above the Secret level and is not approved for storage of any sensitive compartmented information. Thus, it is not available for storage of the majority of the relevant Intelligence Community information, including the information on Hazmi and Mihdhar. In the absence of effective methods for recording and retrieving information obtained from other intelligence agencies, the benefits of increased information sharing among the agencies will remain of limited use. Based on the system in effect during this period, the value of the information was minimal, unless the information was relayed to an individual who could immediately use the information or the information related to an ongoing FBI investigation. When, as here, subsequent additional information increases the significance of the prior information, the absence of an effective information retrieval system effectively precludes any meaningful effort by the FBI to analyze the disparate pieces of information over time. In sum, despite the fact that some personnel at the FBI were aware in January 2000 that Mihdhar was possibly linked to al Qaeda operations and traveled to Malaysia to meet with other suspected al Qaeda operatives, this information was unavailable for further analysis or use once the SIOC closed down in late January or early February 2000. Because no one was assigned to document, follow up, or track the information on Mihdhar, the FBI’s opportunity to discover Mihdhar’s valid U.S. visa during this period and therefore try to locate him was lost. 4. Lack of appropriate infrastructure in FBI field offices Information sharing with the FBI also was impeded by the inadequate facilities for the handling of intelligence information in the two field offices 328 most directly involved in the Hazmi/Mihdhar matter. Intelligence information from the CIA is often classified at a high level. As a result, safeguards must be taken in handling the information, while still allowing appropriate FBI employees the ability to access and use the information. Unfortunately, the FBI’s field offices generally lacked both the necessary physical infrastructure and information technology to readily use this type of information. Without the appropriate physical infrastructure, the FBI will not be able to handle sensitive information in an effective manner. To handle SCI classified material, employees must store and review such information in a SCIF. Access to the SCIF is limited to individuals with the appropriate clearance level and the need to know the information in the SCIF. Adequate security measures must be implemented to prevent unauthorized individuals from gaining access to the spaces containing such materials. The type of equipment that may be brought into the space is also strictly limited. For example, cellular telephones, two-way pagers, and other unsecured communication devices are prohibited. Telephones in SCIFs must be designated for secure transmissions. Computer networks also must be secured for transmission of information. During our review, we observed the workspaces in the FBI New York and San Diego Field Offices and found that they were not set up to adequately handle the type of information involved in the Hazmi and Mihdhar cases. These workspaces were not adequately secured to permit FBI personnel to handle CIA and NSA information at their own desks, even if they had been given the information. Nor were the SCIFs suitable to permit agents to regularly access or handle such information. In the New York Field Office, for example, the SCIF we were shown was extremely small. The CIA detailee to the JTTF worked in this SCIF, but there was little room for any other personnel to enter, let alone use it as a workspace. In the San Diego Field Office, a small SCIF was used as a secure communications center for the entire office. The San Diego office lacked a separate SCIF for the JTTF,264 including the CIA 264 We were informed that a separate SCIF for the JTTF is under construction in the San Diego Field Office. However, this SCIF will only be large enough to accommodate three or four employees at any one time. 329 representative assigned to the task force. As a result, the San Diego agents were hampered in their ability to access CIA information. We also found that New York and San Diego FBI agents did not have sufficient access to secure telephones, known as Secure Telephone Unit third generation or STU III telephones. The limited STU III phones available had to be shared among numerous agents. Again, this made communications involving classified material within the FBI or with other members of the Intelligence Community more difficult. An entire squad comprising as many as 25 individuals shared one or two STU III phones. In addition, as noted above, the FBI agents did not have access to computer systems that could store much of the information received from the CIA. The computers at each agent’s desk in the New York and San Diego Field Offices only provided access to ACS. This system does not permit storage or access to any information classified above the Secret level or any information deemed sensitive compartmented information. Therefore, even if the FBI recipients of the CIA information regarding Hazmi and Mihdhar had wanted to document and store such information in a retrievable fashion, they could not have stored it on the system that FBI agents use. The FBI had no internal system in New York and San Diego that allowed them to use the type of information involved in the Hazmi and Mihdhar case. In addition, most FBI agents in the field did not have direct access to CTLink, the shared Intelligence Community database that did contain some of the information on Hazmi and Mihdhar, such as the NSA information. Field agents could not access, let alone conduct research, on this system. As a result, even if the New York and San Diego agents wanted to search for relevant information about Hazmi and Mihdhar, any sensitive or highly classified information obtained from the NSA and CIA could not be stored in the one system that they used. In contrast, we observed that the CIA’s workspaces permitted their employees to access highly classified information on computers in their personal workstations. Each CIA employee had their own secure computer on which they could receive and research highly classified material. They had several secure telephones that could be used to discuss Top Secret information with others. The difference in CIA and FBI workspaces was particularly stark in the FBI’s San Diego Field Office where, due to the lack of access to an 330 appropriate SCIF, the CIA employee co-located with the FBI’s San Diego Field Office could not access CIA systems. To access CIA systems, he had to travel to a domestic CIA station. 5. OIG conclusion on impediments to information sharing In sum, significant and systemic problems that were evident in the FBI’s handling of the Hazmi and Mihdhar case inhibited information sharing between the FBI and CIA. The FBI failed to define the roles and responsibilities of the FBI detailees to the CTC’s Bin Laden Unit. The FBI failed to ensure effective oversight of the detailees at the CTC. The FBI and the CIA failed to develop a clear understanding of the function of detailees from each other’s agencies. The FBI failed to understand the CIA’s reporting process. The FBI lacked an adequate computer system and appropriate infrastructure for handling intelligence information not directly related to a specific investigation. Although these systemic problems affected the flow of information between the FBI and CIA, we do not believe they fully explain the FBI’s failure to obtain the critical information on Hazmi and Mihdhar. Employees at both the CIA and the FBI failed to provide or seek important information about Hazmi and Mihdhar, despite numerous interactions between them on issues related to Hazmi and Mihdhar from January 2000 through August 2001. We found these interactions were substantive and that much of the information about Mihdhar and Hazmi was exchanged through these ongoing efforts. Unfortunately, the critical pieces of information relating to Hazmi and Mihdhar did not become known to the FBI until shortly prior to September 11. As a former CTC Bin Laden Unit Deputy Chief aptly summarized it to us, “information that should have been shared was not, repeatedly.” B. The actions of the San Diego FBI In addition to issues that affected information sharing between the FBI and the CIA, the FBI had other opportunities to find information about Hazmi and Mihdhar before the September 11 attacks. The time that Hazmi and Mihdhar spent in San Diego was an opportunity during which the FBI could have obtained information about them but did not. As discussed above, Hazmi and Mihdhar entered the United States in January 2000 and moved to San Diego in February 2000, where they resided unbeknownst to the FBI. While in San Diego, Hazmi and Mihdhar associated with Omar al-Bayoumi, a person 331 whom the FBI had previously investigated, and they also lived with an active, FBI informational asset. Yet, the FBI did not become aware of their presence in San Diego until after September 11, 2001. Because Bayoumi spent a significant amount of time with Hazmi and Mihdhar in early 2000, it is possible that – had a full field investigation of Bayoumi been open at the time – the FBI could have discovered Mihdhar and Hazmi’s presence in San Diego and also uncovered the CIA information about their attendance at the Malaysia meetings. Because Hazmi and Mihdhar lived with an FBI asset, it is also possible that if the FBI had documented their presence in San Diego, it would have provided additional investigative leads that could have aided the New York FBI in locating them in August 2001. We therefore evaluated the San Diego FBI’s investigation of Bayoumi and the decision to close its preliminary inquiry on him in June 1999. We also examined the San Diego FBI control agent’s decision not to obtain or document information from his information asset about Hazmi and Mihdhar, who were boarders in the asset’s house. In examining the San Diego Field Office’s handling of the Bayoumi investigation and the informational asset, we also found that, despite the fact that FBI Headquarters had established counterterrorism as a top priority of the FBI in 1998, the San Diego Field Office was continuing to pursue drug trafficking as its top priority in 2001. While the FBI made counterterrorism its top priority on paper, the FBI took few steps to ensure that field offices complied with this directive. We discuss this issue at the end of this section. 1. The San Diego FBI’s preliminary investigation of Bayoumi As discussed above, Bayoumi is a Saudi national who in January 2000 had been living in the United States for approximately six years, was well-paid by a Saudi company that contracted with the Saudi government, and was involved in setting up mosques in the San Diego area. Hazmi and Mihdhar met Bayoumi in Los Angeles approximately two weeks after entering the United States in January 2000. A few days later they moved to San Diego, where Bayoumi assisted them in obtaining an apartment in the complex where he lived. They lived in this complex for four months. Bayoumi’s name had first surfaced at the FBI in 1995 in connection with other investigations. Bayoumi’s name resurfaced at the FBI on August 31, 332 1998, when his apartment manager contacted the FBI to report her suspicions regarding Bayoumi’s activities. The manager reported that she had been notified by the U.S. Postal Inspection Service in March 1998 that Bayoumi had been sent a “suspicious” package from the Middle East. According to the manager, the package had broken open and had a number of wires protruding from it. She reported further that the apartment complex maintenance man had noticed a number of wires protruding beneath the bathroom sink in Bayoumi’s master bedroom. She reported that there had been large meetings of men, who based upon their dress appeared to be Middle Eastern, gathering in Bayoumi’s apartment on weekend evenings. She also complained that several parking spots were being illegally used by the people gathering at Bayoumi’s apartment. On September 8, 1998, the San Diego FBI opened a preliminary inquiry on Bayoumi.265 The assigned agent checked FBI indices for further information regarding Bayoumi and conducted other investigative steps. The agent contacted the U.S. Postal Inspection Service in reference to the alleged “suspicious” package sent to Bayoumi. A postal inspector advised the FBI agent that “suspicious” did not necessarily mean “nefarious,” and the vast majority of suspicious packages were benign. The postal inspector reviewed the report relating to the Bayoumi package and told the agent that the package had been deemed “suspicious” because it had no customs papers or appropriate postage and originated in Saudi Arabia. According to the report, there was no record of any wires protruding from the package, Bayoumi had retrieved the package, and it was no longer called a “suspect parcel.” According to the FBI agent, the apartment manager agreed to record the license plate numbers of the meeting participants. However, the manager later advised the agent that meetings had dwindled to a few participants and then stopped all together. 265 In accordance with the Attorney General’s Foreign Counterintelligence Guidelines, a preliminary inquiry could be opened when there was information or allegations indicating that an individual is or may have been an international terrorist or a recruitment target of an international terrorist organization. Preliminary inquiries were permitted to remain open for 120 days and had to be closed unless the FBI obtained sufficient evidence to open a full field investigation. 333 The agent asked fellow FBI agents to ask their “logical sources” for information regarding Bayoumi. The sources related the following concerning Bayoumi: • Bayoumi was married with small children and had recently completed a master’s degree program and he was looking for a Ph. D. program, but his test scores were too low. He was approximately 30 years old and unemployed. • Bayoumi was a Saudi who regularly attended the ICSD (Islamic Center of San Diego). He was married with children and was working on a master’s or other advanced degree. • Bayoumi reportedly delivered $400,000 to the Islamic Kurdish community in El Cajon, California in order to build a mosque. Source opined Bayoumi “must be an agent of a foreign power or an agent of Saudi Arabia.” • Bayoumi was in the U.S. on a student visa but was applying for a green card. Bayoumi claimed to have a master’s degree and was working on a Ph. D. His father was sending him $3,000 a month for support while he was in school. The FBI agent also contacted the INS in reference to Bayoumi’s immigration status. An INS special agent advised that Bayoumi was in the U.S. on an F-1 student visa, but his work visa had expired. However, the INS reported that his visa could be renewed. The FBI agent received no further substantive information in response to various information checks. According to the agent, the only remaining option was to conduct an interview of Bayoumi. After her supervisor consulted with fellow FBI agents who were working on a large, sensitive counterterrorism investigation involving an alleged terrorist organization, the supervisor instructed the agent not to conduct the subject interview of Bayoumi.266 The agent told the OIG that she did not believe the decision was inappropriate 266 The file indicates that the decision not to conduct an interview was due to an investigation that included a proposed proactive element. The FBI believed that the benefits of interviewing Bayoumi did not justify the risk to the proposed operation. 334 based on the potential effect of such an interview on the other sensitive investigation. On June 7, 1999, the FBI closed its preliminary inquiry on Bayoumi, and he was no longer actively under investigation by the FBI. The FBI case agent told the OIG that she had no concrete information linking Bayoumi to any terrorist activities. She stated that the allegations that gave rise to the preliminary investigation were not substantiated. With respect to the source reporting that Bayoumi had received large sums of money from overseas, the case agent explained it was not unusual for foreign students, especially from Saudi Arabia, to regularly receive money, even large sums of money. Therefore, the case agent did not consider this to be inherently suspicious. The agent’s squad supervisor at the time and other agents on the squad also told the OIG that it was not unusual or suspicious for Saudi students to have received large sums of money from Saudi Arabia. As stated above, one source had provided unverified information that Bayoumi could potentially be a Saudi intelligence operative or source. According to the agent, Bayoumi was allegedly very involved and interested in Saudi affairs in San Diego, and this probably led to the suspicions about Bayoumi’s connection to the Saudi government. However, the agent told the OIG that Saudi Arabia was not listed as a threat country and the Saudis were considered allies of the United States. 267 Therefore, Bayoumi’s potential involvement with the Saudi Arabian government would not have affected the FBI’s decision to close the preliminary inquiry. The squad supervisor at the time of our investigation, who had been an agent on the squad for several years, told the OIG that before September 11, the Saudi Arabian government was considered an ally of the United States and that a report of an individual being an agent of the Saudi government would not have been considered a priority. Other agents on the squad also said that a source reporting that an individual was an agent of the Saudi government 267 Country threats are defined by the FBI as foreign governments or entities whose intelligence activities are so hostile, or of such concern, to the national security of the United States that counterintelligence or monitoring activities directed against such countries are warranted. 335 would not have been cause for concern because the Saudi government was considered an ally of the United States. In addition, the case agent explained that more intrusive investigative techniques could not be conducted because of the restrictions of the Attorney General FCI Guidelines in effect at the time. No meaningful surveillance could be conducted, no bank records or other financial records could be sought, and very little investigative activity beyond fully identifying the individual could be done. In sum, we do not believe that the FBI’s actions with regard to Bayoumi and its decision to close the preliminary inquiry were inappropriate. The agent conducted logical investigative steps that were permitted under the Attorney General Guidelines in effect at the time, such as checking FBI records for information, asking other intelligence agencies for information about the subject, and asking agents to query their sources about the subject, but the agent did not uncover any information to support the allegations. The Guidelines did not permit the case agent to engage in more intrusive investigative techniques, such as a clandestine search of Bayoumi’s property, obtaining his telephone or financial records, or secretly recording his conversations. Although the Attorney General Guidelines would have permitted a subject interview of Bayoumi prior to closing the preliminary inquiry, the decision not to conduct an interview appeared warranted, given its possible effect on an ongoing significant investigation. 2. The FBI’s handling of the informational asset As described above, in May 2000 Hazmi and Mihdhar began renting a room in the home of an FBI informational asset. An FBI San Diego Special Agent who we call “Stan” was the asset’s control agent since the asset was opened in 1994. The asset had provided the FBI with significant information over the years and was considered a reliable source. He was well known in the Muslim community. He often rented rooms in his house to Muslim men in the community who needed temporary housing. At the time that Hazmi and Mihdhar moved in with him, he had two other individuals renting rooms in his house. Mihdhar lived with the asset until June 10, 2000, when he left the 336 United States, and Hazmi remained as a boarder at the asset’s home until December 2000. According to Stan, the asset told Stan that two young Saudis who had recently come to the United States to visit and study had moved in as boarders. The asset described them as good Muslims who often went to the mosque and prayed. The asset provided Stan with their first names but little other identifying information. Stan did not obtain any additional information from the asset about the boarders, such as their last names, and he did not conduct any investigation of them. Had Stan pursued information about Hazmi and Mihdhar, he might have uncovered the CIA information about them. In addition, he might have created a record in FBI computer systems about Hazmi and Mihdhar’s presence in San Diego, which would have provided the FBI with additional information and avenues of investigation when it began to search for them in August 2001. For these reasons, we examined Stan’s actions with regard to the asset. In interviews with the JICI staff and in congressional testimony, Stan stated that the informational asset primarily provided information about the activities and identities of persons in the Muslim community in San Diego who were the subjects of FBI preliminary inquiries or full field investigations. 268 Stan said that the asset volunteered some information about other individuals as well. He said he thought that the asset had good judgment about which individuals might pose a threat and that his reporting had been “consistent” over the years. We reviewed the asset’s file and noted the asset provided information on a regular basis on a variety of different individuals and topics. Although we could not evaluate the asset’s judgment from the file, we consider Stan’s description of the asset’s reporting to be apt. Stan also stated that he was aware that the asset had boarders in his house over the years, and the fact that two new boarders had moved in with the asset did not arouse suspicion. He noted that the asset volunteered that the two boarders were living with him soon after they moved in, but the asset provided the information about his boarders as part of a personal conversation and not 268 As noted above, Stan has retired from the FBI and declined to be interviewed by the OIG. 337 because the asset believed that it had any significance. Stan stated the information provided from the asset was that the two boarders were from Saudi Arabia, which, according to Stan, was not a country that the United States had placed on the list as a threat to national security. Stan said that the asset did not describe his boarders as suspicious or otherwise worthy of further scrutiny. He also asserted that he was prohibited from further pursuing the information about Hazmi and Mihdhar, including documenting the information that he had obtained, because of the Attorney General Guidelines in effect at the time. In examining Stan’s actions, we first considered whether the Attorney General’s FCI Guidelines were applicable to the situation involving Hazmi and Mihdhar. As suggested by Stan, the Attorney General’s FCI Guidelines were designed to ensure that the FBI opened preliminary inquiries and conducted investigations only if the required predicating information was present. Because there were no allegations or information provided to Stan that Hazmi and Mihdhar were terrorists or agents of a foreign power, we agree that Stan did not have sufficient information to open a preliminary inquiry and actively investigate Hazmi and Mihdhar. We also considered whether, at a minimum, Stan could have attempted to obtain additional information about people who were living with his informational asset, such as their full names, and whether he was required to document the information on Hazmi and Mihdhar that he had received from his asset. First, we reviewed FBI policies and procedures for handling assets. Those policies did not require Stan to obtain information from an informational asset about people living in the asset’s house or to conduct record checks to obtain this information. In addition, the policies do not appear to require Stan to have documented information received from the asset about anyone living with him, or to even document their full identities if he had obtained that information. We also interviewed several FBI agents who were on Stan’s counterterrorism squad and asked them whether it would have been their practice to seek additional information about boarders living with an informational asset and what, if anything, they would have done with this information. We found no consensus among them about whether information on boarders like Hazmi and Mihdhar who lived with an informational asset should have been obtained and documented. Some agents stated that they would have pursued more information about boarders living with an 338 informational asset, while others stated that they would not have. Some of the agents stated that they would have noted the fact of the informational asset having boarders in his file. Some agents stated that they would have documented the identities of the roommates in an EC that would have been uploaded to ACS. However, former San Diego Division Special Agent in Charge William Gore told the OIG that he “did not believe anything had been done wrong” in the handling of the informational asset and he did not fault Stan for not obtaining the information. While we recognize that no FBI policy addressed this issue and there was a lack of consensus on what should have been done in a situation like this, we believe that it would have been a better practice for Stan to have questioned the informational asset about his boarders and obtained their full identities. Stan was aware that Hazmi and Mihdhar were relative strangers to the informational asset, and that they were not friends, family, or long-time associates of the asset. Stan also was aware that the asset had no direct knowledge of Hazmi and Mihdhar’s backgrounds and could not vouch for their character. Moreover, the boarders in the asset’s home were in a position to put the asset and the information he supplied to the FBI in jeopardy. Therefore, prudence and operational security would suggest that information about persons living with the asset should have been sought, at least to the extent of learning and documenting their names, and perhaps running a records check on them. If Stan had asked more questions about the asset’s boarders, he also may have acquired enough information to pursue further inquiry. For example, the asset has stated after the September 11 attacks that Hazmi and Mihdhar did not make telephone calls from his house, and that in retrospect he found this behavior to be suspicious. The asset also stated after September 11 that he had told Hazmi to stay away from Bayoumi because of his alleged association with the Saudi government. Therefore, if Stan had asked the asset a few more questions about Hazmi and Mihdhar and acquired this kind of information, it may have led Stan to conduct further inquiries, particularly since Bayoumi had been the subject of an FBI investigation. Moreover, while no specific FBI policy required agents to obtain information about persons living in a house with an informational asset, FBI policies required control agents to continuously evaluate the credibility of their informational assets. Before informational assets are approved, they are required to undergo a background investigation to assess their suitability, 339 credibility, and “bona fides.”269 Certain minimum checks were required, such as a check of FBI indices, local criminal checks, and CIA traces. The policy provided that additional checks “may be deemed necessary,” such as querying other assets and running indices checks on immediate family members. In addition, FBI policy provided that an asset’s bona fides “should be continually addressed,” even after the initial assessment was completed. More specifically, the FBI field office is required to conduct a yearly evaluation of each informational asset and provide the evaluation report to FBI Headquarters. This report is required to contain, among other things, the FBI’s number of contacts with the informational asset during the reporting period, a summary of the most significant information furnished by the informational asset, the number of preliminary inquiries and full investigations that were opened based on information provided by the informational asset, and “steps that have been taken to establish asset bona fides since last evaluation.” Although Stan would not have been required to obtain additional information about his informational asset’s boarders to complete this report, the FBI’s policy of continually vetting the credibility of its assets permitted Stan to seek more information about Hazmi and Mihdhar and the other boarders from his asset and run indices checks on any persons living with his informational asset. We reviewed the informational asset’s file, Stan’s yearly evaluation of the asset, and Stan’s reporting on the bona fides checks conducted on the informational asset. Based on our review, we were concerned by the lack of information included in the file in support of the bona fides checks conducted by Stan each year. In each of the documents provided to FBI Headquarters about the informational asset that we reviewed, Stan wrote the following perfunctory paragraph: “Asset bona fides have been established through independently received reliable asset reporting, [redacted] and physical surveillance.” Stan maintained no predicating information in the file on these bona fides checks. The file did not disclose which checks or surveillance had been 269 The FBI defines “bona fides” to mean that the asset or informational asset “is who he/she says he/she is;” that the asset “has the position or access the asset claims to have;” and that the asset “is not working for or reporting to a foreign intelligence service or international terrorist organization without the knowledge of the FBI.” 340 conducted, by whom, when, or the results. Without that material, the informational asset’s bona fides were merely verified through the attestation of Stan. It is possible that Stan conducted numerous indices checks and conducted an exhaustive bona fides check on the informational asset each year. It also is possible that he conducted minimal or no checks and merely attested to the informational asset’s credibility based on their personal history and relationship. Because we were unable to interview Stan, we could not determine which was more likely. However, no FBI policy described the level of detail to be contained in an asset file. We believe the policy should require an asset file to contain at least minimal information to allow a reviewer to independently verify that an adequate background check has been conducted. This information is necessary to allow FBI managers to determine whether the control agent is continuing to assess each informational asset’s credibility. This information would also help ensure that the control agent has not become too comfortable with the informational asset and thus vulnerable to being misled or failing to obtain adequate information about the asset. We also were concerned by the lack of policy or practice specifying what information from the asset must be documented. The Hazmi and Mihdhar case clearly demonstrates that information must be documented to be useful. Even if Stan had obtained the full names of Hazmi and Mihdhar from the informational asset, he would not have been required to document it in any retrievable format. Without the requirement to document such information, the information would not have been accessible to other FBI personnel. For information to be useful, it must be documented in a retrievable form and it must be available for consideration and analysis. In sum, we believe that Hazmi and Mihdhar’s presence in San Diego should have drawn some scrutiny from the FBI. Although unknown at the time, documenting their presence in San Diego in a searchable and retrievable manner would have provided an opportunity for the FBI to connect information in the future. If Hazmi and Mihdhar’s presence in San Diego in 2000 had been documented, an FBI indices record check in August 2001, when the FBI received information from the CIA that Hazmi and Mihdhar had entered the 341 United States, might have led the FBI to the San Diego information. This connection would have provided substantive leads for the New York FBI’s effort to locate Mihdhar in August 2001.270 3. San Diego FBI’s failure to prioritize counterterrorism investigations As discussed in Chapter Two, in 1998 the FBI adopted a 5-year strategic plan that established the FBI investigative priorities in a 3-tier system. Tier I priorities were “foreign intelligence, terrorist, and criminal activities that directly threaten the National or Economic Security of the United States.” Tier II priorities were “crimes that affect the public safety or undermine the integrity of American society: drugs, organized crime, civil rights, and public corruption.” Tier III priorities were “crimes that affect individuals and property such as violent crime, car theft, and telemarketing scams…” On March 15, 1999, shortly after Director of Central Intelligence George Tenet asserted the U.S. Intelligence Community was declaring war on Usama Bin Laden and al Qaeda, FBI Headquarters established national level priorities within its Counterterrorism Program. Bin Laden and al Qaeda, along with the Bin Laden-allied Egyptian Islamic Jihad (EIJ) and al Gama’at al Islamiyya (IG), were designated as “priority group one” for the FBI’s counterterrorism efforts. Our review of the Hazmi/Mihdhar chronology revealed no appreciable shift in resources by the FBI’s San Diego Field Office in response to these changed priorities. We found that prior to September 11, 2001, the actual investigative priority for the San Diego Field Office was drug trafficking. According to former San Diego Special Agent in Charge William Gore, the highest concentration of FBI agents and resources in San Diego was directed at combating drug trafficking based on the FBI’s process and procedures used each year to set priorities in its field offices. He said that white-collar crime was the office’s second priority, and violent crime was its third priority. 270 As noted, Mihdhar and Hazmi used their own names to open bank accounts, conduct financial transactions, obtain state identification cards, purchase a vehicle, obtain telephone service, take flying lessons, and rent an apartment while residing in San Diego. 342 Counterterrorism was only the fourth priority for the San Diego FBI office. The counterterrorism efforts in San Diego were directed primarily at another terrorist organization and related groups not connected to Al Qaeda, and the majority of San Diego’s counterterrorism investigations targeted activities related to the indirect support of terrorism conducted by those groups. We found that the San Diego FBI focused little to no investigative activity on al Qaeda prior to September 11. San Diego FBI personnel stated to us that they had believed there was no significant al Qaeda activity in San Diego based on information from their sources and investigative activities. The former supervisor of the San Diego counterterrorism squad explained their job at the field office level was to “shake the tree and see what fell out” in relation to potential terrorism activities in their area. Although San Diego agents assigned to counterterrorism conceded they had received little to no specific training concerning Bin Laden or al Qaeda, they asserted that al Qaeda did not have a significant presence in San Diego prior to September 11. Yet, al Qaeda was present in San Diego, unbeknownst to the FBI. Hazmi and Mihdhar resided in San Diego. Unfortunately, the San Diego agents were not focusing on al Qaeda. Even though FBI Headquarters had designated al Qaeda as the number one counterterrorism priority, the San Diego FBI was not attempting to identify individuals that were associated with al Qaeda. Since September 11, many San Diego agents have been moved from other squads and assigned to counterterrorism. Significantly, the San Diego office opened a large number of intelligence investigations on potential al Qaeda subjects immediately after September 11. Obviously, the focus and priorities dramatically changed after September 11. But there is no reason to believe the al Qaeda presence in San Diego began only after September 11. If San Diego’s focus on counterterrorism and al Qaeda had occurred earlier in San Diego, there would have been a greater possibility, though no guarantee, that Hazmi’s and Mihdhar’s presence in San Diego may have come to the attention of the FBI before September 11. However, it is important to note that San Diego’s allocation of resources before September 11 and the lower priority it gave to the Counterterrorism Program were not atypical of FBI field offices before September 11. In an OIG September 2002 audit report entitled “A Review of the Federal Bureau of Investigation’s Counterterrorism Program: Threat Assessment, Strategic 343 Planning, and Resource Management,” we found that “Although the FBI has developed an elaborate, multi-layered strategic planning system over the past decade, the system has not adequately established priorities or effectively allocated resources to the Counterterrorism Program.” Furthermore, the OIG report found that resources were not allocated consistent with the FBI’s priorities – particularly at the field office level – because of the lack of “management controls” in the FBI’s “complicated and paper-intensive strategic planning process.” Instead of allocating resources based on FBI priorities, field offices allocated resources primarily based on previous caseloads in the field office. According to the report, prior to September 11, “the Bureau devoted significantly more special agent resources to traditional law enforcement activities such as white collar crime, organized crime, drug, and violent crime investigations than to domestic and international terrorism investigations.” For example, in 2000 twice as many FBI agents were assigned to drug enforcement than to counterterrorism. Thus, the San Diego’s office allocation of resources was not different from many other FBI field offices, despite the stated priorities of the FBI. C. Events in the spring and summer of 2001 As described in the factual chronology, the FBI had several opportunities in the spring and summer of 2001 to obtain critical intelligence about Mihdhar and Hazmi. Although the FBI and the CIA were discussing Mihdhar, Khallad, and the Cole investigation throughout the spring and summer of 2001, the FBI did not become aware of the critical intelligence involving Mihdhar’s U.S. visa and subsequent travel to the U.S. until late August 2001. As we discussed above, we believe that systemic problems regarding information sharing between the two agencies contributed to the FBI’s failure to obtain this information earlier. But restrictions within the FBI also contributed to the FBI’s failure to acquire critical information about Hazmi and Mihdhar before September 11. In this section, we discuss those problems. 1. Restrictions on the flow of information within the FBI By the summer of 2001, the effect of the various restrictions within the FBI on information sharing – commonly referred to as “the wall” – had resulted in a nearly complete separation of intelligence and criminal investigations within the FBI. This separation greatly hampered the flow of 344 information between FBI personnel working criminal and intelligence investigations, including information concerning Hazmi and Mihdhar in the summer of 2001. As discussed in Chapter Two, in late 1999 the FISA Court had become the “wall” for purposes of passing FISA information on targets of a particular terrorist organization from FBI intelligence investigations to criminal investigations. Any information that intelligence agents wanted to give to criminal agents had to be provided to the FBI’s NSLU, which then provided it to OIPR, which then provided it to the FISA Court, which then had to approve the passage of the information to criminal agents. In addition, after the FISA Court was notified in the fall 2000 about errors in approximately 100 FISA applications, a significant portion of which related to the FBI’s representations about the “wall” procedures in al Qaeda cases, the FISA Court imposed new restrictions on the FBI’s handling of FISA information. The FISA Court required a certification from all individuals who received FISA information stating that they understood this requirement. The FISA Court exempted CIA and NSA personnel, who often received FISA information from the FBI, from this certification requirement. But the FISA Court required that the CIA and NSA indicate on the information they provided to the FBI whether the information had been obtained based on FISA information previously provided to them by the FBI (called “FISA-derived information”). In response, the NSA decided that it was more efficient not to delay dissemination of intelligence while checking to see if it was derived from FISA, and it therefore placed a caveat on all NSA counterterrorism reports to the FBI stating that before information could be considered for dissemination to criminal personnel, the FBI had to check with the NSA General Counsel about whether the intelligence was FISA-derived. Once the NSA determined whether the information was FISA-derived, the FBI had to comply with the wall procedures for passing FISA-derived information to criminal agents or prosecutors. If the information was not FISA-derived, it could be passed directly. FBI Headquarters personnel became wary that any involvement of criminal agents in intelligence investigations could present problems for the FBI with the FISA Court. A former ITOS unit chief described the FISA Court’s certification requirement as a “contempt letter” and said that it “shut down” the flow of information in the FBI. He further stated that FBI 345 Headquarters employees became worried that any misstep in handling FISA information could result in harm to their careers because an FBI agent was banned from appearing before the FISA Court and OPR began an investigation on him. These three factors – the Court had become the screener in al Qaeda cases, the certification requirement imposed by the FISA Court, and concerns about violating the Court’s rules – combined to stifle the flow of intelligence information within the FBI. FBI employees described this to the OIG as the walls within the FBI becoming “higher” over time. New York FBI agents told the OIG that the walls were viewed as a “maze” that no one really understood or could easily navigate. As we discuss below, these walls affected the FBI personnel’s discussions about the Mihdhar information at the June 11, 2001, meeting in New York and the FBI’s decision to open an investigation to locate Mihdhar in August 2001. 2. Problems at the June 11 meeting At the June 11, 2001, meeting, FBI Headquarters and CIA CTC personnel discussed with New York FBI investigators issues relating to the Cole investigation. At the time of this meeting, the FBI analyst who we call Donna had received information from the CIA concerning travel in January 2000 of an al Qaeda operative named Khalid al-Mihdhar to Malaysia through Dubai. Donna also had received surveillance photographs from the CIA showing Mihdhar meeting with other unidentified al Qaeda operatives in Malaysia.271 After receiving the information from the CIA, Donna had conducted her own record check on Mihdhar in CTLink and discovered the NSA information from late 1999 and early 2000 associating Mihdhar with a suspected terrorist facility in the Middle East linked to al Qaeda activities and his plans to travel to Malaysia in January 2000. 271 Although not shared with Donna or known to anyone else in the FBI, the CIA also knew in June 2001 that Mihdhar had a U.S. visa, that Mihdhar’s associate -- Hazmi -- had traveled to the United States in January 2000, and that the Cole mastermind Khallad had been identified in one of the Kuala Lumpur surveillance photographs. 346 This NSA intelligence about Mihdhar would have been important to the FBI agents conducting a criminal investigation of the Cole attacks. However, Donna did not share this information with the criminal agents at the June 11 meeting because of concerns about the wall. By this time, the FBI was operating under the requirement that all NSA counterterrorism information had to be reviewed by the NSA’s General Counsel’s Office for a determination of whether it was FISA-derived before it could be considered for dissemination to criminal agents. Because she had not yet asked the NSA whether the information could be passed, Donna did not provide the New York agents with any of the NSA information. That information would have been important to the New York agents who were working the Cole investigation because they specialized in al Qaeda operations and at the June 11 meeting showed great interest in the Malaysia meetings and Mihdhar. That information may also have provided the criminal agents with additional leads and could have led to the information that Mihdhar and Hazmi had traveled to the United States in January 2000. We recognize that the caveat on sharing any NSA counterterrorism information did not mean that the criminal agents were prohibited from ever obtaining access to the NSA information on Mihdhar. But if the information was FISA-derived, the caveat created a delay in the criminal agents receiving the information because of the lengthy procedures that had to be followed to share the information with them. With respect to the information Donna had received from the CIA about the Malaysia meetings, Donna showed the photographs to New York agents and asked whether they could identify Cole participant Fahd al Quso in the photographs. After one of the agents made a tentative identification, the agents asked questions about Mihdhar and the photographs. The agents continued to ask Donna questions about Mihdhar, the Malaysia meetings, and the photographs on June 11 after the meeting. As we discussed above, it is unclear how much questioning occurred during the actual meeting and how much occurred after the meeting. Donna was unable to answer most of the agents’ questions because she had not obtained the information from the CIA. This, in our view, was not because of the wall, but was because of Donna’s failure to plan the meeting adequately or ask sufficient questions from the CIA in advance of the meeting. 347 First, we believe the planning for the June 11 meeting was flawed. Although Donna and other IOSs frequently traveled to New York to work on the Cole investigation, she told the OIG that this was the first time that she had arranged for a meeting involving CTC personnel in New York. Yet, according to what the meeting participants told the OIG, the purpose and the agenda of the meeting were not clear. The participants agreed that they knew there was going to be a discussion of the investigative results on the Cole attack. The New York agents believed that the CTC and FBI Headquarters had information to share with New York. Donna and the CTC participants, however, described the meeting as a “brainstorming” session to determine what new leads could be pursued and what FBI Headquarters could do to assist New York. No agenda was prepared and no supervisors were consulted for their input about the meeting. Even though Donna said that she called the meeting to explore further leads or avenues of investigation in the Cole case, she apparently did not ask the CTC participants to be prepared to present information or answer questions. Mary and Peter told the OIG they were not in a position to discuss the Cole investigation. Mary said she was not up to speed about the Cole investigation or the Malaysia meetings. Peter told the OIG that as an analyst at the CIA, he did not have authority to discuss CIA information at the meeting and he was merely “tagging along.” Donna told the OIG that she considered Mary to be another FBI employee at the meeting, and for this reason did not provide her with any specific instructions in preparation for the meeting. Donna also said that she had not invited Peter and because she was not in his chain of command, she did not ask him to be prepared. However, the New York agents we interviewed told the OIG that they believed that CTC personnel were coming to the meeting in part to share information with them. The fact that all the participants we interviewed described the meeting as unproductive and a “waste of time” highlighted that a more useful exchange of information could have occurred. With respect to the Kuala Lumpur photographs, Donna had obtained only limited information from CIA employee John about the photos when she received them. She did not ask general background questions such as whether anyone else in the photographs had been identified, or what else was known from the Malaysia meetings. Donna told the OIG that because she believed the CIA provided her with everything she was entitled to know, she did not have 348 an in-depth discussion about the photographs. John said he did not recall anything about his discussions with Donna regarding the Kuala Lumpur photographs. Donna told the OIG that when the New York agents asked her questions about Mihdhar, the Malaysia meetings, and the photographs, she thought that they were reasonable questions, but she did not know the answers. She stated that at the time she obtained the Kuala Lumpur photographs from the CIA, she believed that they were only potentially related to Quso and their significance to the Cole would hinge on whether Quso was in the photographs. We believe Donna should have asked the CIA additional questions about the photographs. She had reason to believe Quso, a key individual in the Cole investigation, may have attended the Malaysia meetings. Given her interest in whether Quso had attended the meetings, she should have wanted to ascertain, and asked the CIA, what, if anything, was known about the purpose of the Malaysia meetings, who were the other participants at the meetings, what was known about the participants, and any other available information. Donna also did not ask the CIA whether there were additional photos or documentation. Donna told the OIG she was unaware that there could have been additional photographs or other relevant information available. We believe that someone in her position should have known or at least asked for additional information about the subject of the photographs in preparation for the meeting. We also were troubled by Donna’s inadequate efforts to obtain additional information after the June 11 meeting, particularly information about the Malaysia meetings, since it had been the subject of a dispute between Donna and Scott. Although Donna told the New York agents that she would check with the CIA about additional information regarding the photographs and the Malaysia meetings, Donna made little effort to obtain this information until two months later, in August 2001. Donna told the OIG that she believes that she made some unsuccessful follow-up phone calls to Peter and John about the photographs. It is not clear from the documentary evidence how much Donna did before August to obtain the information, but she did not provide additional information to the New York agents about the photographs for at least two months. We recognize that FBI analysts were overwhelmed with assignments and had to juggle many responsibilities, however, given the possible 349 connections of this information to the Cole investigation, we believe Donna should have made more aggressive and timely efforts to obtain this information soon after the June 11 meeting and to keep the New York agents informed about what her follow-up efforts were. By the same token, Scott, the New York Cole case agent, did little to follow up after the June 11 meeting to obtain information he requested about the Malaysia meeting. Scott told the OIG he “often” asked Donna about the status of the information, but he was not provided any such information. Donna contended that Scott did not follow up on his June 11 requests. We found no evidence such as e-mails or other documents to support Scott’s claim that he raised the issue often with Donna. We believe that neither Donna nor Scott made significant efforts after the meeting to obtain the information. 3. The FBI’s investigation in August 2001 to find Mihdhar and Hazmi As discussed above, on August 22, 2001, the FBI learned that Mihdhar and Hazmi had entered the United States in January 2000, that Mihdhar had again flown to New York on July 4, 2001, and that there was no record of either of them leaving the country. The FBI also learned that Khallad had been identified in the Kuala Lumpur photographs. Upon discovery of this information, the FBI opened an intelligence investigation in New York in an effort to locate Mihdhar. Once again, however, the separation between intelligence and criminal information affected who could receive access to the information about Hazmi and Mihdhar. This interpretation of the wall also hampered the ability of the FBI New York agents working on the Cole investigation to participate in the search for Hazmi and Mihdhar. In addition, we found that the FBI’s efforts to locate Hazmi and Mihdhar were not extensive. We do not fault the case agent assigned to locate them. He was new and not instructed to give the case any priority. Rather, we found that the FBI New York did not pursue this as an urgent matter or assign many resources to it. 350 a. The effect of the wall on the FBI’s attempts to locate Mihdhar As discussed above, Donna drafted an EC to the New York FBI requesting it open an investigation to locate Mihdhar. She also called Chad, the FBI New York agent who primarily handled intelligence investigations for the Bin Laden squad, to give him a “heads up” about the matter, and she subsequently sent the EC to him. She wrote in the e-mail that she wanted to get the intelligence investigation going and the EC could not be shared with any of the agents working the Cole criminal case. Chad forwarded the EC to his squad supervisor, Jason, who nevertheless disseminated the EC via e-mail within the Bin Laden squad, including to the criminal agents assigned to the Cole investigation. Scott read the EC and contacted Donna regarding it. Donna informed Scott that he was not supposed to have read the EC because it contained NSA information that had not been cleared to be passed to criminal agents. Donna told Scott that he needed to destroy his copy. Scott responded that the effort to locate Mihdhar should be part of the Cole criminal investigation, and he argued with Donna regarding the designation of the investigation as an intelligence matter. Donna asserted that, because of the wall, criminal agents were not yet entitled to the underlying intelligence provided by the NSA, and without that predicating material, the FBI could not establish any connection between Mihdhar and the Cole criminal investigation. Scott, Donna, and acting UBL Unit Chief Rob then spoke via conference call. Scott argued that the investigation should be opened as a criminal investigation and that more resources and agents could be assigned to a criminal investigation by New York. He also argued that criminal investigative tools, such as grand jury subpoenas, were far quicker in obtaining information than the tools available in intelligence investigations. Donna consulted with an NSLU attorney, Susan. According to Donna, Susan concurred that the matter should be handled as an intelligence investigation and that because of the wall, a criminal agent could not 351 participate in the search for or any interview of Mihdhar.272 When Donna advised Scott of Susan’s opinion in an e-mail message, Scott responded by email that he believed the wall was inapplicable. Scott ended his message by suggesting that because of the NSLU’s position, people were going to die and that he hoped that NSLU would stand by its position then. The way that FBI Headquarters handled the Mihdhar information reflected its interpretation of the requirements of the wall prior to September 11. First, because the predication for the search for Mihdhar originated from the NSA reports, this information could not be immediately shared with criminal agents. Instead, it first had to be cleared for dissemination by the NSA, which would determine whether the intelligence was based on FISA information. If so, the information had to be cleared for passage to the criminal agents – the information had to be provided to the NSLU, which then provided the information to OIPR, which then provided it to the FISA Court, which then had to approve the passage of this information to criminal agents. In fact, the limited INS information concerning Mihdhar’s and Hazmi’s entries into the United States was the only unrestricted information in the EC immediately available to the criminal investigators. As in the Moussaoui case, the decision to open an intelligence investigation resulted in certain restrictions. FBI Headquarters employees understood that they needed to ensure that they avoided any activities that the FISA Court or OIPR could later deem “too criminal” and could use as a basis to deny a FISA application. This included preventing a criminal agent from participating in a subject interview in an intelligence investigation. While Scott was correct that the wall had been created to deal with the handling of only FISA information and that there was no legal barrier to a criminal agent being present for an interview with Mihdhar if it occurred in the intelligence investigation, FBI Headquarters and NSLU believed that the original wall had been extended by the FISA Court and OIPR to cover such an interview. Scott’s frustration over the wall was similar to Henry’s in the Moussaoui investigation, when Henry was told by Don that seeking prosecutor 272 As discussed above, Susan told the OIG that she did not recall this discussion with Donna. 352 involvement prematurely could potentially harm any FISA request. Scott, like Henry, wanted to pursue a criminal investigation and became frustrated when he was advised by FBI Headquarters that he could not proceed in the manner he deemed appropriate. Scott’s perception was that FBI Headquarters had misconstrued “the wall” and the wall had been inappropriately expanded. He told the OIG that he believed the wall should only relate to FISA or FISAderived information. Like the Minneapolis FBI, Scott believed that he was being “handcuffed” in the performance of his job and that FBI Headquarters “erred on the side of caution” in its approach to intelligence information. FBI Headquarters, on the other hand, acted in accordance with its experience with OIPR and the FISA Court. FBI Headquarters believed that OIPR and the FISA Court required strict adherence to the procedures for the passage of intelligence information to criminal investigations and required separating criminal and intelligence investigations. Donna explained that the FISA Court’s mandates resulted in the need for the FBI to create a near complete separation between intelligence and criminal investigations in order to effectively use intelligence information. Rob also told the OIG that there were “land mines” in dealing with intelligence versus criminal information, and it was difficult to appropriately straddle the two sides. Our review of this case showed that the wall had been expanded to create a system that was complex and had made it increasingly difficult to effectively use intelligence information within the FBI. The wall – or “maze of walls” as one witness described it – significantly slowed the flow of intelligence information to criminal investigations. The unintended consequence of the wall was to hamper the FBI’s ability to conduct effective counterterrorism investigations because the FBI’s efforts were sharply divided in two, and only one side had immediate and complete access to the available information. The wall was not, however, the only impediment in the FBI’s handling of the investigation to find Mihdhar and Hazmi. We found there were also other problems in how the search for Mihdhar and Hazmi was handled. b. Allocation of investigative resources We found that prior to the September 11 attacks, the New York Field Office focused its al Qaeda counterterrorism efforts on criminal investigations, but it did not expend a similar effort on intelligence investigations or the 353 development of intelligence information. New York agents told the OIG they believed that criminal prosecution was the most effective tool in combating terrorism. They asserted that criminal investigations are also a preventive activity and the FBI had always focused on preventing terrorism, even before September 11. They pointed to the TERRSTOP investigation in 1993, an investigation to uncover a terrorist plot to attack New York City landmarks, and the criminal investigation into the East African embassy bombings. Prosecutors also argued that criminal investigations and prosecutions are an effective preventive measure against terrorism. Testifying before the Joint Intelligence Committee, Mary Jo White, the former U.S. Attorney for the Southern District of New York (SDNY), stated, “[W]e viewed the terrorist investigations and prosecutions we did from 1993-2002 as a prevention tool.” Patrick Fitzgerald, currently the U.S. Attorney for the Northern District of Illinois and formerly an Assistant U.S. Attorney in the SDNY, told us that it is a misconception that there has to be a difference between prosecution and gathering intelligence. He added that the SDNY prosecutions produced a “treasure trove of [intelligence] information.” However, prosecutors also realized criminal investigation and prosecution were not the only means of countering terrorism. White stated, “the counterterrorism strategy of our country in the 1990s was not, as I have read in the media, criminal prosecutions.” She further stated, “none of us considered prosecutions to be the country’s counterterrorism strategy, or even a major part of it.” As Fitzgerald told us, “in order to connect the dots, you need people to gather the dots.” Although we agree criminal investigations are a highly effective counterterrorism tool, intelligence investigations were not given nearly the same level of resources and attention in the FBI’s New York Field Office before September 11, 2001. This criminal focus was clear in the assignment of personnel on the New York Bin Laden squad. From October 2000 to June 2001, only one agent on the Bin Laden squad was designated as the “intelligence” agent – the agent we call “Chad.” The remainder were designated as “criminal” agents.273 Chad told us that he was inundated with 273 One criminal agent worked on intelligence matters on a part-time basis. 354 intelligence investigations and information, and he rarely had enough time even to review all the incoming Bin Laden intelligence information, let alone to digest, analyze, or initiate the procedures to pass the information to the criminal agents where applicable. Chad also told us that the “intelligence” agent designation was “not a desirable position” within the Bin Laden squad. He described himself as the “leper” on the squad due to “the wall.” Furthermore, Chad stated that the intelligence side of the squad received far less and lower quality resources. The handling of the investigation to locate Mihdhar provides a clear indication of the primacy of the criminal over intelligence investigations in the New York office. On August 28, 2001, the New York Field Office opened an intelligence investigation to locate Mihdhar based upon Donna’s EC. Donna told the OIG that she believed there was some urgency to the Mihdhar investigation, not because of any evidence that he was operational, but because he could leave the United States at any time and the opportunity to find out as much as possible about him would be lost. She said she therefore called Chad about the EC in advance, which she did not normally do. However, when she sent the EC to New York, she assigned the matter “routine” precedence, the lowest precedence level. When asked about this discrepancy, Donna told the OIG that the Mihdhar investigation was “no bigger” than any other intelligence investigation that the FBI was pursuing at the time. The New York Bin Laden squad relief supervisors, who we call “Jay” and “David,” told the OIG that they recognized that there was some urgency to the Mihdhar investigation. Yet, the FBI in New York did not treat it like an urgent matter. The investigation was given to an inexperienced agent – ”Richard” – who had only recently been assigned to the Bin Laden squad. This was his first intelligence investigation. As one of the largest field offices in the FBI, with over 300 agents assigned to the JTTF, the New York Field Office could have assigned additional or more experienced agents who were not involved in the Cole criminal investigation to assist Richard. However, the New York Field Office Bin Laden Squad was focused on criminal investigations. As a result, the designation of the Mihdhar matter as an intelligence investigation, as opposed to a criminal investigation, undermined the priority of any effort to locate Mihdhar. 355 Finally, we also noted that there was a clear predicate for a criminal investigation that no one appeared to notice at the time. In her EC, Donna noted that Mihdhar had previously traveled to the United States, according to information she had obtained from the INS. After the FBI’s intelligence investigation was opened, she obtained and forwarded to Richard a copy of Mihdhar’s June 2001 visa application on which he stated that he had not previously been issued a visa and had never traveled to the United States. Thus, there was a clear basis to charge Mihdhar criminally with false statements or visa fraud. Significantly, this information had been provided to the FBI without the restrictive caveats placed on NSA reports and other intelligence information. As a result, if Mihdhar had been found, he could have been arrested and charged with a criminal violation based on the false statements on his visa application. However, the FBI did not seem to notice this when deciding whether to use criminal or intelligence resources to locate Mihdhar. D. Individual performance This section summarizes the performance of individual FBI employees in the Hazmi and Mihdhar matter. While none of them committed misconduct, we believe that several FBI employees did not perform their duties as well as they could have and should have. We address in turn the FBI employees involved in each of the five lost opportunities. In this section, we do not discuss the performance of individual CIA employees. However, we believe that a significant cause of the failures in the sharing of information regarding the Hazmi and Mihdhar case is attributable to the actions of the CIA employees. It is the responsibility of the CIA OIG to assess the accountability of the actions of CIA employees. 1. Dwight In January 2000, intelligence information was developed about Hazmi, Mihdhar, and other al Qaeda operatives meeting in Malaysia. Dwight, an FBI detailee to the CTC’s Bin Laden Unit, read the CIA cables about the Malaysia meeting. The cables indicated that Mihdhar had a U.S. visa and that he listed New York on the visa application as his intended destination. Dwight recognized the significance of this information to the FBI and drafted a CIR to pass this information to the FBI. 356 Unfortunately, his draft CIR was never sent. A notation added to the CIR suggested that it was held at the request of the CIA’s Deputy Chief of the Bin Laden Unit. Several FBI detailees accessed the CIR, and Dwight inquired about it again five days later, asking the Deputy Chief in an e-mail whether it was going to be sent or whether he needed to “remake” it in some way. We found no response to his e-mail, and none of the participants, including Dwight and the Deputy Chief, said they remembered this CIR at all. We believe the primary responsibility for the failure to pass this information rests with the CIA. The evidence indicates that the CIA did not provide permission for the CIR to be sent.274 However, we also believe that Dwight should have followed up as much as necessary to ensure that the information was sent to the FBI. Although we found evidence that he inquired once about the disposition of the CIR, we found no additional evidence that he continued to follow up to ensure that the information was sent. If Dwight was stymied in his attempt to learn about the disposition of the cable, or if the CIA gave no reasonable explanation for why the information was not being sent, he could have brought this issue to the attention of another supervisor in the CTC. In our view, Dwight took the commendable initiative to draft the CIR to share the information with the FBI, but did not follow through adequately to ensure that it was sent, and the information in the CIR was not provided to the FBI until shortly before the September 11 attacks. 2. Malcolm Malcolm was a New York FBI agent detailed for several years to the CTC. He told the OIG that he understood his role at the CTC was, among other things, to be the “eyes and ears” of the New York Field Office. We do not believe that he performed this role sufficiently. He acknowledged to the OIG that one of his duties was “to monitor” New York Field Office cases, but he said he read only the cables that he thought were “interesting,” generally 274 The CIA has asserted that the information in the CIR was sent to the FBI through another cable, which may be why the CIR was not sent. A CIA cable stated that Mihdhar’s travel documents, including a multiple entry U.S. visa, had been copied and passed “to the FBI for further investigation.” As discussed above, however, we found no evidence that this cable was correct and that this information had actually been provided to the FBI. 357 based solely on his review of the cable subject line. In addition, while he said his role was to “facilitate inquiries of mutual interest,” the only example he could provide was his acting as a liaison for FBI offices around the country by following up on tracing requests and reporting on their status. This was not very onerous or substantive. We believe that FBI management is primarily responsible for failing to provide the FBI detailees to the CTC, including Malcolm, with clear duties, direction, and supervision. But we believe Malcolm should have done more and taken more initiative in performing his duties at the CTC. 3. Stan For several months in 2000, Hazmi and Mihdhar lived as boarders in the house of an FBI informational asset. The asset briefly mentioned the two boarders to his FBI control agent, who we call “Stan.” Stan did not document this information, seek to learn the boarders’ full identities, or conduct any checks on them. No FBI policy required Stan to seek or document this type of information from the asset, and we found differences among the other FBI agents who we interviewed about whether they would have sought such information from an asset. While Stan did not violate any specific FBI policy, we believe it would have been a better and more prudent practice for him to have sought at least minimal information from his asset about the boarders living with him. The asset knew little about the boarders, and the boarders could have compromised information provided by the asset to the FBI. Moreover, FBI policy required Stan to continually evaluate the asset’s credibility and provide a yearly evaluation report on the asset. Stan’s yearly report on this asset was minimal, with a bare attestation of the asset’s bona fides. It contained no indication of what evidence Stan had used to make these attestations. While we do not suggest that Stan had to conduct extensive reviews of everyone living with the asset, Stan’s actions in following up on this information were not particularly thorough or aggressive. 4. Max In January 2001, a joint FBI/CIA source identified Khallad in photographs of the Malaysia meetings. Because the FBI ALAT who was 358 involved in the handling of the source, Max, was unable to speak any of the joint source’s languages, a CIA employee conducted the debriefings of the source, including the debriefing in which the source identified Khallad. We concluded that Max was not informed of the source’s identification of Khallad from the Kuala Lumpur surveillance photograph, either at the time of the identification or afterwards. Although CIA cables covering the debriefing described the identification of Khallad, these were not shared with Max. Instead, he saw CIA TDs that did not contain the information about the identification. CIA documents do not indicate that the ALAT was informed of the identification, and no other evidence indicates that the ALAT knew. We found that the ALAT included detailed descriptions in his reports of other information from the source, which indicates he was not provided the information about the identification of Khallad. We also found that the New York FBI agents who interviewed the source in February 2001 were not informed of the identification of Khallad. In sum, we believe the ALAT did not learn about the source’s identification, not that he knew about identification but failed to share this information with others. We believe that, as the ALAT, Max should have been more familiar with the CIA’s reporting process. He was not aware that the CIA’s TDs contained only a part of the information obtained during the source debriefings. Although our review revealed that many FBI employees operated with misunderstandings about the ways the CIA recorded and reported intelligence information, a significant function of the ALAT position is to interact with the CIA. Had he recognized that he could not rely on TDs for full reporting about the source’s information, he could have asked his CIA counterpart directly for any additional information from the source, and the ALAT may have learned about the identification of Khallad. In addition, given Max’s concern that he provide FBI Headquarters with all of the information reported by the source, it would have been prudent for him to consult with the CIA case officer and ask sufficient questions to ensure that he had received all of the information. We found no indication that he did so. 5. Donna Donna, the FBI analyst who worked on the investigation of the Cole attacks, planned a June 11, 2001, meeting with the Cole investigators and CIA 359 employees to discuss information relating to the Cole investigation. She deserves credit for organizing this meeting and seeking to share intelligence information with the Cole investigators. However, we fault her performance in two respects. First, we found that the meeting was poorly planned, and Donna did not clearly communicate the purpose of the meeting to the participants. Donna also failed to obtain significant information prior to the meeting that could have been shared with the investigators about the Malaysia meetings. After the meeting, although Donna devoted a significant amount of time to the Cole investigations, she did little specific follow-up to provide answers to the investigators about their logical questions regarding the Malaysia meetings. We believe she did not do all she could have to acquire that information for the New York agents, even though she had said that she would as a result of their discussion at the June 11 meeting. As a result, the FBI missed another opportunity to focus on Mihdhar and Hazmi earlier than it did. When Donna finally learned from Mary on August 22, 2001, that Hazmi and Mihdhar were in the United States, Donna quickly and appropriately took steps to have the FBI open an investigation to locate them. She personally called the New York Bin Laden intelligence agent and told him about the matter. This was an unusual step to call the agent directly, and it suggested that the investigation should be given some priority. However, when she sent the EC to New York, she designated the EC as having a routine precedence. Donna’s actions indicated some urgency in the need for the investigation yet the subsequent EC did not convey any urgency. The New York Field Office assigned the case immediately, and the agent began working on the case within two business days of the assignment. If the EC had conveyed urgency, the FBI New York Field Office might have assigned additional or more experienced agents to locate Mihdhar and Hazmi and initiated the search sooner. 6. Rob We believe that Rob, as Donna’s supervisor, is also responsible for Donna’s failures. While the FBI at the time permitted IOSs to make significant decisions, often with little supervisory input, we believe that as a supervisor, he should have ensured that she was handling the June 11 meeting appropriately and, if necessary, become involved with the planning or execution of the meeting. Although Donna often traveled to New York to work on the Cole investigation, the June 11 meeting involved the CIA and an AUSA, which 360 should have led to more supervisory involvement in the purpose, agenda, and outcome of the meeting. But Rob had little supervisory involvement with it, either before or after the meeting. In addition, although Donna drafted the EC requesting the investigation of Mihdhar, the EC was ultimately approved and sent by Rob. Therefore, we believe he also bears some responsibility for failing to ensure that the appropriate precedence level was used on the EC. 7. Richard We do not fault Richard for his limited investigation, which was still in the nascent stages by the time of the September 11 attacks. As we described above, Richard took logical steps to try to locate Mihdhar and Hazmi, such as completing a lookout for Mihdhar with the INS, requesting local criminal history checks, checking with New York hotels about Hazmi and Mihdhar, and conducting commercial database checks on them. However, there were many more investigative steps that could have been pursued, in New York and elsewhere, had the investigation been assigned greater priority and had the FBI provided more resources to this investigation. The FBI was not close to locating Hazmi and Mihdhar when they participated in the September 11 attacks. We believe that the FBI in New York should have assigned the matter more priority than it did. 8. Mary Mary was assigned by her CIA managers in May 2001 with finding and reviewing the CIA cables relating to the Malaysia meetings and their potential connection to the Cole attack. Mary did not find the relevant CIA cable traffic until late July and mid-August 2001. She told the OIG that she did not have time to focus on this assignment until then. Upon discovering on August 21 that Hazmi and Mihdhar had traveled to the United States, she immediately passed this information to the FBI. We recognize that the disparate pieces of information about the Malaysia meetings were not easy to connect and that the task of developing patterns from seemingly unrelated information was complex. Yet we question the amount of time that elapsed between Mary’s assignment and her discovery of the important information. As we discussed previously, however, Mary’s assignments were directed and controlled by her managers in the CTC. We, therefore, leave this issue to the CIA OIG for its consideration. 361