« PrécédentContinuer »
U.S. DEPARTMENT OF JUSTICE, FEDERAL BUREAU OF INVESA 10. 1994.
Washington, DC, May 10, 1994. Hon. PATRICK J. LEAHY, Chairman, Subcommittee on Technology and the Law, U.S. Senate, Washington, DC.
DEAR PAT: Enclosed please find our responses to the questions both you and Sen. ator Pressler had regarding the Administration's proposed “Digital Telephony and Communications Privacy Act of 1994" which were forwarded to me in your two letters dated March 24, 1994. Pursuant to discussions with Subcommittee staff, specific information concerning the Bureau's informal survey which identified ninetyone instances where law enforcement's electronic surveillance were frustrated due to advanced telecommunications technologies will be forthcoming under separate cover.
Should you have any questions concerning this or any other issue of mutual concern, please do not hesitate to contact me. Sincerely yours,
LOUIS J. FREEH
LOUIS J. FREEH'S RESPONSES TO QUESTIONS SUBMITTED BY SENATOR PRESSLER Question. Why is further legislation necessary? What is lost now that will be gained by this legislation?
Answer. There is no disagreement that under current law telecommunications companies and others are mandated to assist law enforcement agencies execute court orders for electronic surveillance and pen registers/trap and traces, pursuant to court orders directing them to do so. The issue is the nature and extent of that mandate. The current language of the "assistance provision" in the various federal electronic surveillance and pen register/trap and trace statutes, though unambiguous in its mandate, is silent on (and therefore gives no guidance concerning) what level of effort or level of cost, if any, must be expended by a provider when its syg. tem, as deployed or planned, does not readily accommodate execution of electronic surveillance or pen register/trap and trace court orders, or, when impediments are encountered, how quickly an effort must be undertaken and completed.
Attorneys for the FBI, Bellcore, and the telecommunications industry believe the current language does not clearly define the responsibilities of service providers, particularly when some level of effort is required beyond the rudimentary furnishing of line information and leased line facilities. The current language was adequate when enacted in 1970, since telephony technology then was relatively simple and there were relatively few service providers. Today's telecommunications marketplace is vastly different, more complex, and abounding with service providers (e.g., over 2,000 common carriers alone).
Hence, it is imperative that the generic electronic surveillance requirements of law enforcement be articulated formally, that common carrier responsibilities be spelled out, and that a reasonable compliance period be set. Legislation is the only way to put all common carriers on notice regarding these requirements and to legally obligate the over 2,000 carriers to act in a timely and comprehensive fashion. Further, legislation is needed in order to formally enlist the cooperation of support service providers and equipment manufacturers upon whom common carriers rely for provisioning service. So also it is necessary to define the government's legal and equitable remedies regarding common carriers who fail or refuse to comply or who are dilatory, and to provide a mechanism for reimbursement to common carriers for costs expended in making the modifications associated with achieving compliance.
If there were any doubt about whether new legislative language is needed to clarify and memorialize the above-mentioned requirements and responsibilities, one need only consider the complaints and false accusations made by some opposition spokespersons concerning the proposed legislation's memorialization of law enforcement's long-standing requirements.
A detailed statement of the inadequacy of the current "assistance provision" language is set forth at pages 24-27 of Director Freeh's statement submitted for the record.
Question. How much of a problem is this today? In how many actual cases, roughly, has the FBI been unable to accomplish a court-approved wiretap due to a telecommunications company's inability to make the tap?
Answer. The problem, though hard to quantify, is significant and is clearly growing. The informal FBI survey documented 91 such problems, but there have been many more. Over the last decade, we estimate that several hundred court orders have been frustrated or were not sought because of these problems.
With the introduction and growth of cellular telephone systems, and the introduction and use of sophisticated features and services, the number of court orders for electronic surveillance and/or pen register/trap and traces that were frustrated, in whole or in part, or delayed, or were not sought, grew. With the on-going introduction of even newer, more sophisticated features and services, wireless systems, etc., the number of technological problems will grow at an accelerated rate and the ability of law enforcement agencies to fully and properly execute these court orders will continue to erode.
There is no single repository for information regarding the number of court orders that could not be properly executed with regard to the hundreds of criminal Title III wiretap and thousands of pen register court orders executed each year by hundreds of law enforcement agencies throughout the U.S. it was precisely this circumstance that led the FBI to conduct an informal survey in 1992 to attempt to gather some data, however imprecise, regarding this serious problem. This informal survey disclosed 91 "recent" problems. Information regarding these problems was retrieved by FBI field office Technically Trained Agents (TTAS), and most of the information reported on pertained to FBI cases. It is unclear how effective these TTAS were in obtaining sensitive information from state and local law enforcement authorities about problems they were encountering in executing court orders, particularly regarding then pending electronic surveillance efforts.
The FBI relies upon Technically Trained Special Agents (TTA) throughout the United States to technically implement the execution of the court orders. This duty also involves technical liaison with the telecommunication companies. Problems encountered by these TTAs are sometimes brought to the attention of FBIHQ through requests from the field TTAs for assistance from the FBIHQ's Engineering Section. It is usually through these requests that such problems become known, though they are not tabulated. In addition to the court orders that cannot be executed (fully executed), many requests for electronic surveillance or pen registers/trap and traces by field office Special Agents have been discouraged by the TTAs in a number of field offices due to known on-going impediments in a particular area that have frustrated court orders in the recent past. Thus, a significant number of requests were never pursued and never reached the stage of being submitted for court approval. Such "discouraged" requests are not reported formally to FBIHQ and they were not counted in the survey. A third category of problems are those where court orders have been delayed due to capacity limitations within the telecommunications companies' switches. Although such court orders may ultimately be executed, they may be delayed to the point where critical and valuable information is lost or not collected. Finally, there have been a great number of court orders that are not successfully executed because of problems that do not permit the comprehensive interception and acquisition of the subject's communications and/or "call "setup information" (dialing information), and therefore have been of limited investigative value. See pages 29 35 of Director Freeh's statement for the record regarding this issue.
Question. Is the problem that the technology is too sophisticated or is it that the companies do not provide adequate access points?
Answer. Generally, the latter. The level of sophistication of the new technology is not viewed as a problem. To the contrary, the sophisticated switches and equipment being developed and deployed today are more capable than the less sophisti. cated switches that have been in use for many years. As service providers deploy more sophisticated switches, they develop more sophisticated features and services in response to or in anticipation of the commercial demands of their customers. If service providers view the electronic surveillance requirements of law enforcement as sophisticated functional "features," there is no reason to believe that the switches (which are essentially specialized computers) cannot accommodate this functionality. The real problem is that the companies that manufacture and operate this equipment currently have no economic incentive or clear, specific legal requirement to include these functions as features in their developmental effort. If the needs of law enforcement agencies are accommodated during the developmental phase of new systems, features, and equipment, as is the case with other commercial services and features that are marketed to the public, adequate capabilities and access for electronic surveillance can be provided to law enforcement agencies without any adverse impact on the networks and without delaying the introduction of new commercial services and features.
Question. Can you describe several examples?
Answer. During the early 1980's, there was an explosive growth in the use of cellular telephones. The need to conduct electronic surveillance and pen register investigations within cellular systems became evident at an early stage. Unfortunately, the systems had been developed several years earlier and were being deployed rapidly. After unsatisfactory attempts to intercept cellular telephones on the radio frequency portion of the cellular system, law enforcement agencies sought the assistance of cellular providers in order to execute the court orders. Many cellular service providers (and there are literally hundreds of new companies for law enforcement agencies to deal with) were not aware of, or always responsive to, their legal obligation to assist law enforcement agencies with the execution of court orders. After a number of years, a number of technical solutions were eventually developed and deployed on a case-by-case basis (often with the government funding the system upgrade). The delayed deployment of these solutions led to an unknown (but we estimate significant) number of court orders not being executed, not being fully executed, or being delayed in their execution. Additionally, since these solutions were not considered in the development cycle of the cellular equipment and were of a "band aid," retrofit nature, in many instances the capacity range of some of these solutions was limited, which affected the number of simultaneous court orders that could be conducted, and in others their overall capability or functionality was lim
Another example is call forwarding. When call forwarding was first introduced by telecommunications providers, the element that "forwarded" Or re-directed telephone calls was the central office switch. As such, even though a subject might forward his calls across the country, the content of the call itself still passed through the central office switch that served his telephone instrument. However, once forwarded, the subject's call did not appear in the local loop law enforcement's tradi. tional access point. Further, since no provision was made for lawful law enforcement electronic surveillance access during the development and deployment of call forwarding, service providers themselves were unable to afford law enforcement an effective means of acquiring such forwarded communications, even though they pass through their switches.
Question. I understand that the General Services Administration (GSA) was opposed several years ago to an earlier FBI proposal regarding this sage issue. Are they opposed to this latest version? Please explain what changes were made from the earlier version to the latest one?
Answer. As you are aware, the Administration's Digital Telephony legislative proposal, as well as an earlier legislative proposal formally transmitted to the 102nd Congress under the Bush Administration, deals with the issue of law enforcement's continued technical ability to conduct court-authorized electronic surveillances as advanced telecommunication technologies, services, and features are introduced into the nation's telecommunications networks. GSA is not a law enforcement agency and therefore has no expertise in the area of court-authorized electronic surveil. lances, nor does it have any particular expertise relative to any possible impact the incorporation of law enforcement's electronic surveillance requirements might have with regard to current or future networks, systems, features, or services. (By comparison, neither the FCC nor the National Institute of Standards and Technology (NIST), which also possesses telecommunications expertise, interposed any objection to the legislative proposal.) Unfortunately, in 1992, GSA comments were solicited by OMB in the context of an expedited legislative proposal clearance process, without GSA being privy to the numerous, lengthy, and in-depth legal and technical discussions and briefings which preceded the clearance process.
Consequently, although GSA's comments were considered along with those from other agencies within the Bush Administration, given the above, GSA's comments were evaluated accordingly. The important position was not GSA's but that of the Bush Administration. The Bush Administration determined that the only realistic way to effectively address the "digital telephony problem was through a comprehensive legislative solution. Such a comprehensive legislative solution was formally transmitted to the 102nd Congress by the Bush Administration in September of 1992; however, time did not permit any action being taken on it.
In April, 1993, President Clinton directed that an interagency working group be established to examine, among other things, telecommunication technology trends and their impact on court-ordered electronic surveillances. After an in-depth eight month study, the Clinton Administration likewise determined that the only effective way to address the so-called "digital telephony" problem is through a comprehensive legislative solution; and a formal administration legislative proposal, entitled the "Digital Telephony and Communication Privacy Improvement Act of 1994,” was transmitted to Congress on March 25, 1994. Fundamental to this decision was the Clinton Administration's belief, like that of the Bush Administration, that it would
be unacceptable for the safety of the American public to be imperiled, the national security to be endangered, and effective law enforcement to be eroded through the diminishment or loss of this critical and essential tool of our nation's law enforcement and intelligence agencies.
The current legislative proposal is a significant improvement over the prior proposal, and includes a number of new and important provisions, many of which were developed after we received input from the telecommunications industry regarding the 1992 proposal. The current legislative proposal focuses on where the problems are within the networks of common carriers. Hence, all other types of service providers (computer networks, PBX operators, etc.) have been eliminated from coverage. The time period for common carrier compliance has been doubled from 18 months to three years. The wording of the electronic surveillance requirements has been refined based on industry suggestions. A "systems security provision” has been added to reflect that any Premises/switch-based intercepts will only be initiated by designated common carrier personnel. The assistance of the FCC is included. Importantly, a provision for on-going "consultation” using industry standards bodies and other fora is established to ensure cost-effective compliance, and the Attorney General is authorized to reimburse common carriers for all reasonable costs directly associated with the modifications required to be performed in order for common carriers to comply with the Act, up to $500 million under the current proposal. Finally, the proposed legislation includes provisions that improve communications privacy protection for cordless telephones, certain radio-based data communications, and communications transmitted using certain privacy-enhancing modulation techniques.
Question. The telecommunications industry claims this legislation would go well beyond the status quo by allowing the FBI to be part of the front-end design process, where it has not been previously. How do you respond to their concerns?
Answer. The industry's claim is erroneous. The purpose and effect of the Administration's proposed legislation is to clarify and define the existing responsibilities of common carriers to provide the "technical assistance necessary to accomplish the interception.” The proposed legislation does not require common carriers to design their systems in any one, particular way; and it certainly doesn't contemplate the FBI or any other law enforcement agency dictating network design. The legislative proposal simply requires that common carriers be able to respond to the long-standing generic electronic surveillance requirements of law enforcement and be able to meet their current legal responsibility to effectively “assist” law enforcement when served with a court order.
This legislation also does not propose that industry design systems with the ability to wiretap as the "design goal." Law enforcement is simply requiring common carriers to craft appropriate electronic surveillance solutions that interface with their networks' designs. Industry is in the best position to develop reasonable and cost-effective design solutions which accommodate, rather than frustrate, court orders for electronic surveillance and pen registers/trap and traces, and, at the same time, maintain the security and reliability of telecommunications networks.
Further, this legislation does not prohibit or impede the development or deployment of new telecommunications technologies. We are unaware of any authoritative industry statement that these requirements would significantly delay development or deployment of new technologies. Advanced technologies support a number of features that make systems more intelligent. The basic technology underpinning the introduction of advanced features and services, which often impedes electronic surveillance, likely will also offer solutions to these very problems.
Finally, law enforcement's.requirements will not dictate the course or pace of technology. However, with this legislation, our requirements will intentionally be included, rather than excluded, in the development of new technologies in much the same fashion that other legislation requires new technologies and products to take public safety into account and in much the same way that new commercial features and services are developed and deployed by telephone companies for purely economic gain. (For further information, see pgs. 54-55 of Director Freeh's statement for the record.)
Question. As telecommunications technology progresses will the FBI accidently acquire "content-based” information after only receiving authority to obtain "transactional" information (e.g., banking information) such as with a pen register, which only requires a grand jury subpoena-a lower standard to meet than a warrant issued frog a judge? How should the FBI address this problem? Is "content" information, obtained according to a subpoena, not a warrant, admissible in court? Should there be a heightened standard for content information, somewhere between "probable cause" and "reasonable suspicion"?
Answer. At the outset, the foregoing questions include a number of mistaken assumptions of fact, law, and investigative procedures. By clarifying these, a meaningful response best can be made. First, contrary to the false assertions of some in the privacy community, it should be made absolutely clear that the FBI and other law enforcement agencies do not focus investigative efforts on individuals who are not centrally involved in serious (felony) criminality. Further, United States Attorneys/ District Attorneys do not authorize grand jury subpoenas or court orders, and courts do not issue court orders, where relevancy to a criminal investigation is not demonstrated. There are no “routine surveillances” of innocent people. Nor are court orders or subpoenas issued for irrelevant, noncriminal, or privileged information.
Second, under Federal law (18 U.S.C. 3121–27), a pen register lawfully acquires "numbers dialed or otherwise transmitted on the telephone line to which such device is attached.” 18 U.S.C. 3127(3). Such dialing information or "call setup information" is distinguished under law from call content which is dealt with under the Title III provisions (18 U.S.C. 2510_21) and from telephone transactional records which are dealt with under 18 U.S.C. 2701–11.
Third, use of pen registers by the FBI requires a Federal court order issued from a court of competent jurisdiction. Pen registers My not be used based upon a grand jury subpoena.
Fourth, if Title III content information is introduced in court, the provisions of Title III, including obtaining a Title 111 court order, must be followed in order to introduce such evidence. The use of pen registers to acquire the numbers dialed is expressly excluded from the treatment of interception of communications content under Title III. 18 U.S.C. 2511(h)(i).
Fifth, when the FBI acquires dialed numbers through the use of a pen register, these numbers mean nothing per se to the FBI. To determine the subscriber information pertaining to the party called (through the numbers dialed), the FBI bust then obtain a second form of legal process under 18 U.S.C. 2703 (a subpoena or court order) and serve it on the telephone company. The telephone company in such situations, of course, only possesses and hence only provides the telephone subscriber or listing information that identifies the called party (e.g., (202) 324-3000 is listed to FBIHQ, 10th and Penn. Ave., NW., Washington, DC.).
Sixth, assuming that the telephone number dialed by the criminal subject is determined to be (after subpoenaing the telephone company subscriber records) a banking or credit card institution, and assuming the FBI desires to examine the subject's financial records, then the FBI is legally obligated to obtain a third form of legal process and serve it on the institution, pursuant to the provisions of the Right to Financial Privacy Act, codified at 12 U.S.C. 3401-3422, especially 3413.
To the extent that dialing information (call setup information) tells something about criminal subjects and their activities, this has been the case since the first primitive "pen registers" were employed decades ago. Under the proposed legislation, law enforcement would acquire this dialing information as it does today-no more no less.
Although we are sensitive to legitimate privacy concerns, we have not received any information on this topic to suggest that any new laws are required or that current laws require amendment. For example, the notions propounded by Mr. Berman that either telephone transactional information under 18 U.S.C. 2703, or the call setup information (dialing information) obtained through pen registers under 18 U.S.C. 3121 et seq., or both, should only be available based upon a probable cause warrant is a radical departure from past law and jurisprudence and would cause the acquisition of electronic transactional information or records to be treated vastly differently than documentary transactional information and records. There simply is no legal or rational basis for according any type of telephone transactional information or records higher privacy protection than other documentary transactional information or records. As stated above, the dialing information obtained by a pen register always requires a court order, whereas telephone and documentary transactional information and records can be (and historically always have been) obtained through use of a subpoena.
Whereas the Government's legislative proposal simply seeks to maintain the legal technical status quo, Mr. Berman and the DPSWG appear to be seeking a repeal of the pen register/trap and trace provisions enacted in the Electronic Communications Privacy Act of 1986.
It is particularly puzzling why some privacy rights spokespersons, under the guise of a "false transactional information scare," loudly advocate raising the legal bar to extreme heights for law enforcement's limited access to information concerning criminal subscribers, when telephone companies can afford broad access (and routinely sell, disclose, or exchange) much of the same type of information at issue here with the private sector regarding all subscribers with little or no legal constraint.