[Seat] Summary of informal gathering 05.01.

[Muhammad Usama Sardar <[e-mail]>]

Hi all,

a summary of today's informal gathering follows:

(Attendees: Please feel free to correct me if I have misunderstood 
something or add anything important that I may have missed.)

# *Attendees* (in alphabetic order)

Tiru, Usama

Apologies: Meiling and Yuning

# *Updates*

 1. In addition to the lists mentioned in last two week's summary [0,1]
    (where there is no issued raised), the paper [2] and artifacts [3]
    on vulnerabilities in draft-fossati-tls-attestation have been shared
    since then with wider security communities for their expert opinion:
      * CFRG [4]
      * An easy-to-access list of all communication is maintained at [5].
 2. There was some discussion at WIMSE about attested TLS where they
    use "attestation" in very broad sense.

# *Discussion summary*

We discussed:

 1. Formal modeling and results of different "binding" mechanisms for
    intra-handshake attestation:
     1. Attestation nonce
     2. Early exporter
     3. Hash of server's public key
     4. (Hypothetical) Binding the webPKI cert to the unique platform
        identifier of the Attester: we discussed several issues here.
 2. Formal modeling and results of most popular implementations of
    intra-handshake attestation:
     1. Cocos AI [6]: they are abusing the SNI extension to convey
        attestation nonce.
     2. CCC proof-of-concept [7]
     3. Meta’s AI [8]
     4. Edgeless Systems Contrast [9]: they were abusing the SNI
        extension to convey attestation nonce, and currently abusing the
        ALPN extension to convey attestation nonce.
     5. GA4GH-SDK [10]
 3. Request agenda slots for interim/topics to be covered: (note for
    chairs: this is not an agenda request; this is summary of attendees'
    discussion for preparation for that. We will reach out to you for
    formal requests later after coordination with other authors)
     1. On fairness grounds, I proposed we (proponents of post-handshake
        attestation) limit the sum of our agenda requests to half of
        meeting time in total giving opponents of post-handshake
        attestation/proponents of intra-handshake attestation an equal
        and fair opportunity to present/clarify their perspective.
     2. We will prepare backup slides for detailed discussion for the
        rest of half hour just in case nothing else comes up from the
        opponents.

# *Questions for WG*

 1. Could someone think of any other "binding" mechanisms for
    intra-handshake attestation that is not listed in #1 in "Discussion
    summary" above?
 2. Is someone aware of any other implementation of intra-handshake
    attestation that is missing in #2 in "Discussion summary" above?
 3. Does someone have any opinion on the pending PRs [11,12,13]?
 4. Would any one of you be willing to review our paper on above
    mentioned modeling and give feedback until Jan 9?
 5. Is there any other relevant list (that is missing in [6]) that I
    should reach out to request review of [3] and [4]?
 6. Does someone have any other use case that is not yet covered in use
    cases draft?

*# Important note*

Authors of use cases and expat drafts meet every week according to the 
schedule shared here [14], to which WG participants are very welcome. 
Please note that these informal gatherings are meant to facilitate 
discussion and they have no formal standing in the IETF standards 
process. We will do our best to bring a summary to the WG for 
opinion/discussion.

-Usama

[0] https://mailarchive.ietf.org/arch/msg/seat/cWOxXFIMeHMRMsi1iivjCqR_L3A/

[1] https://mailarchive.ietf.org/arch/msg/seat/0levZNcNWguwueqfIshM0IBqxow/

[2] 
https://www.researchgate.net/publication/398839141_Identity_Crisis_in_Confidential_Computing_Formal_Analysis_of_Attested_TLS

[3] https://github.com/CCC-Attestation/formal-spec-id-crisis

[4] https://mailarchive.ietf.org/arch/msg/cfrg/61SpXcv--YhcLbAvlcRog2J9S5s/

[5] 
https://github.com/CCC-Attestation/formal-spec-id-crisis/wiki/MailingLists

[6] https://docs.cocos.ultraviolet.rs/atls

[7] https://github.com/ccc-attestation/attested-tls-poc

[8] 
https://ai.meta.com/static-resource/private-processing-technical-whitepaper

[9] 
https://github.com/CCC-Attestation/meetings/blob/main/materials/MarkusRudy.contrast-atls-ccc-attestation.pdf

[10] https://github.com/elixir-cloud-aai/ga4gh-sdk

[11] https://github.com/tls-attestation/exported-attestation/pull/47

[12] https://github.com/tls-attestation/exported-attestation/pull/48

[13] https://github.com/tls-attestation/use-cases-and-properties/pull/22

[14] https://github.com/tls-attestation#meetings