Skip to content
Apart Research
Apart Research
Contact us
Sprint projectFeb 2, 2026UK, Germany, Singapore, Australia
4th place

Fingerprinting All AI Cluster I/O Without Mutually Trusted Processors

Naci Cankaya, Jakub Kryś, Jonathan Ng, Luke Marks, Felix Krückel · Team The Bypass Annihilators

Submitted to The Technical AI Governance Challenge. Sprint projects are early-stage work by participants, not Apart Research publications.

Read the report

Report: Fingerprinting All AI Cluster I/O Without Mutually Trusted Processors

Open in a new tab (opens in new tab)
Code (opens in new tab)More on ethernet-frame-visualization.vercel.app (opens in new tab)
Share

We design and simulate a "border patrol" device for generating cryptographic evidence of data traffic entering and leaving an AI cluster, while eliminating the specific analog and steganographic side-channels that post-hoc verification can not close. The device eliminates the need for any mutually trusted logic, while still meeting the security needs of the prover and verifier.

Reviews

Judging this Sprint?

Review this project

Your public critique appears on this page without your name. Your private critique is not published; only the Apart team reads it. If you agree below, we share your review with grantmaking.ai (opens in new tab) and the Transformative AI Fund so strong projects can be funded.

Not shown on this page.

Shown on this page, without your name.

Only the Apart team reads this, and funders if you agree below.

Share my name publicly on grantmaking.ai *
Share my private critique with funders *

How much would this matter for AI safety if it worked? How innovative is it? For scores of 4-5: is this actually new to the field, or replicating recent work?

Scoring guide
  1. 1Negligible. No clear problem addressed, or no meaningful novelty.
  2. 2Limited. Addresses a real problem but with a generic or well-trodden approach. Incremental at best.
  3. 3Moderate. Clear problem with a reasonable approach; some novelty in framing or method beyond routine application of existing tools.
  4. 4Significant. Important problem with an original approach, or identifies a neglected problem area. A valuable contribution others could build on.
  5. 5Exceptional. Tackles a critical AI safety problem with a genuinely novel approach, or opens a new research direction. Clear theory of change. You'd be excited to share this with researchers in the area.

How sound are methodology, implementation, and findings?

Scoring guide
  1. 1Seriously flawed. Methodology broken, results uninterpretable, or implementation doesn't work.
  2. 2Weak. Approach has significant gaps: missing validation, flawed experimental design, or incomplete implementation.
  3. 3Competent. Technically solid given the short duration. Methodology makes sense, results are interpretable, limitations acknowledged, work builds toward clear conclusions.
  4. 4Strong. Thorough methodology with convincing validation. Results clearly support conclusions. Immediately useful for future work.
  5. 5Exceptional. Ambitious scope executed rigorously. Surprising findings, novel methods, or unusually robust validation.

How clearly are work, findings, and impact potential communicated?

Scoring guide
  1. 1Incomprehensible. Cannot determine what the project is actually claiming or doing.
  2. 2Hard to follow. Key information buried, missing, or diluted by excessive length. Significant effort to extract main points.
  3. 3Clear enough. Can understand the problem, approach, and results without undue effort. Core content clearly present: problem, method, findings, limitations.
  4. 4Well presented. Easy to follow, well-structured, appropriate level of detail. Target audience would get it quickly.
  5. 5Exceptionally clear. A pleasure to read. Complex ideas made accessible. Could serve as a model for how to present this type of work.

  1. [Disclaimer: I was not familiar with the OSI model, which the authors say one should be familiar with to fully understand this work].

    Problem really well framed, and interesting as an alternative to approaches relying on trusted hardware. Some sparse and non-central comments (in the interest of constructive criticism on a good paper):

    - the paper’s suggestions for handling encrypted traffic (e.g. key escrows), which is central to the viability of any such scheme, will likely face significant adoption barriers.

    - fact that this seems to largely not apply to distributed training makes it inapplicable to a wide range of potential situations where this kind of mechanism would be useful

  2. I'm super jazzed about this one. It looks really substantial, though a lot of the details are above my paygrade and out of my field (i.e. i didn't like audit it to make sure no mistakes were made). The intersection of steg with certs over egress is I think critical-- most naive egress tricks like with EBPF are not steg-robust. Which I think is an extremely hard (in some sense formally impossible in general) problem, so any progress at all is very helpful and impressive. I'm excited to see more.

Cite this project

@misc{cankaya2026fingerprinting,
  title = {{Fingerprinting All AI Cluster I/O Without Mutually Trusted Processors}},
  author = {Naci Cankaya and Jakub Kryś and Jonathan Ng and Luke Marks and Felix Krückel},
  year = {2026},
  month = feb,
  note = {Submitted to The Technical AI Governance Challenge, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/fingerprinting-all-ai-cluster-io-without-mutually-trusted-processors-d81l}},
  url = {https://apartresearch.com/sprints/projects/fingerprinting-all-ai-cluster-io-without-mutually-trusted-processors-d81l}
}

More from The Technical AI Governance Challenge

All 47 entriesBrowse all projects

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026