Minutes of the April 23-24, 1997, Meeting of the Technical Advisory Committee to Develop a Federal Information Processing Standard for the Federal Key Management Infrastructure A quorum being present, the third meeting of the Committee was called to order at 9:00 a.m. by Executive Secretary Ed Roback. He welcomed everyone and reviewed the agenda for the two day meeting (Reference #1) and asked whether any of the members or federal liaisons had any announcements to make. Mr. Sabo (SSA), briefly provided an overview of SSA's plans to gather expert opinions on on-line availability of SSA information. The WGs being established by SSA have a deadline of June 30th to make their recommendations. Nineteen members were in attendance. This meeting was primarily focused on hearing from the Working Groups as to their activities since the last meeting, and in planning for future work. The WGs are: #1 Framework, #2 Security Models, #3, Key Recovery Agents (KRAs) , #4, non-KRA Elements, and #5, Interoperability. During the two days of discussion, it was agreed that WGs 3 and 4 would be combined. Note that the WGs do not act independently of the Committee. They function solely to gather information or conduct research of the Committee, analyze relevant issues and facts, or draft position papers for consideration by the Committee. Each WG provided an update of their activities; see references. Mr. French of WG #1 (Framework) asked the TAC about what sort of deliverables the TAC expected to produce and what level of detail should be planned for. He also asked for opinions about how the various WG products could be integrated. Mr. French asked each member/liaison to provide him with additional materials, URLs, and so forth that might be of use to the Framework WG. He presented a strawman statement of the deliverables and FIPS outline. He raised questions regarding level of detail, stored vs. transmitted data, access to the key recovery agent(s), the breadth of the FIPS, interoperability issues, and the export issue. Dr. Matyas provided a briefing of two overheads describing a key recovery model. (See references.) This was the subject of significant discussion and questioning, with suggestions made that this model be refined by the WG for discussion at the next meeting. Dr. Brickell of the WG#2 (Security Models) provided an overview of WG #2's current thinking about security models, and identified issues, among others, involving key recovery agents vs. data recovery agents, interoperability, definitions, security, encryption algorithm independence, and owner access. During the discussion, a question was raised regarding what would happen if there is not consensus among the membership on a particular issue. Dr. Kent indicated that when the TAC arrives at a set of requirements that is clearly defined, if there is not consensus on the issue, a vote may be necessary among the members (with appropriate dissenting opinions). Mr. Alexander provided a joint report for WG#3 (Key Recovery Agents) and WG#4 (Non-KRA Elements). It was agreed, because of the very small size of both of these WGs, that they would be combined. (The Secretariat will create a new collective e-mail address.) Mr. Alexander began by stressing that by implementing key recovery, a flaw is being built into the system. He reviewed issues involving, among others, definitions, structure, KRA certification, session vs. master keys, product elements, basic assumptions, and product/system issues. Dr. Kent indicated the need for the WG to see each component in the model be well characterized and then a more detailed requirement developed for the TAC to review. Mr. Paul Clark provided a briefing of WG #5 (Interoperability). (See reference.) He reviewed various modes for key recovery, including decryption and session key recovery. He said that requirements for real-time recovery of encrypted communications have not been adequately defined. He discussed including the originator as a recipient in communications in order to assure recoverability. Mr. Kenneth Thibodeau, Director, Center for Electronic Records, National Archives and Records Administration (NARA) provided the TAC with a brief overview of NARA's activities in the area of electronic recordkeeping. This was in response to requests from TAC members to better understand the government's requirements for retention of records, particularly with regard to the expected duration of the storage. NARA, which is just beginning to address electronic records, takes a very long-term view of storage, using 100 years as a minimum requirement. With respect to encryption, it appears that NARA would prefer that agencies provide their materials to NARA in plaintext, using archiving plans for the Defense Messaging System as an example. (See reference.) During discussion which followed, the Chairman stressed that, while he was generally pleased with WG activities, if the WGs only work on the days before the TAC meeting, the TAC will not be able to complete its taskings. E-mail exchange is a viable way to do this, as are conference calls. He also stressed the need for WGs to produce strawman text (beyond overhead briefings) for comment by TAC members. It was suggested that two weeks before the next (i.e., by June 4), the WGs have materials ready for distribution to the TAC as a whole for review prior to the TAC meeting. No members of the public requested time to address the TAC. All presenters were asked to provide electronic copies of their reports to the secretariat for posting on the TAC's web page. Additionally, any materials (e.g., presentation materials, etc. intended for distribution at the June meetings should be provided to the Secretariat at least ten days before the meeting (June 18-19) for copying. The Secretary discussed plans for TAC meetings in 1998. He intends to distribute a request to TAC members for dates on which they are already busy. After collecting this, he will propose meeting dates. It is envisioned that the TAC will continue to meet six times per year for two-day meetings. In planning for future meetings, the WGs expressed a preference for WG meetings the day before each TAC meeting. Having no further business, Dr. Kent adjourned the meeting at 2:30 p.m. References (on file with the Secretariat): #1 - Agenda #2 - WG #1 (Framework) #3 - WG #2 (Security Models) #4 - WG#3/WG#4 KRA and Non-KRA Elements (Alexander) #5 - WG#3/WG#4 KRA and Non-KRA Elements (Matyas) #6 - WG#5 (Interoperaibilty) #7 - NARA materials ----------------