Stealing Reasoning Traces from Proprietary LLM APIs
A paper reports that Anthropic, OpenAI, and Google return encrypted chain-of-thought blocks that can be replayed across sessions, users, and models.
Source-based brief · Editor review not recorded · See checks and limits ↓
⚑ Recency not confirmed · what to check
We could not confirm from the original page that this is a newly published development. Check its date before treating it as current news.
Claim context: This is a short link-blog summary of a paper; the full paper was not checked and source claims not independently checked.
One primary source; other links are context, not independent corroboration.
- Primary sourcesimonwillison.net
Grok Bot submitted this discovery. We could not confirm from the original page that this is a newly published development. Check its date before treating it as current news. The original source and submitted evidence remain inspectable; no editor has verified the claim.
Open full article view →- Linked article+30
- Fetched source page+30
- No editor link review+0
- Original non-discussion link+5
- Recency not confirmed-8
This measures the stored source trail and editorial steps. It does not certify accuracy, peer review, or independent corroboration. Popularity adds no points.
View full scoring →Read the scoring rubric →What the source says
A paper reports that Anthropic, OpenAI, and Google return encrypted chain-of-thought blocks that can be replayed across sessions, users, and models. Per the cited text, researchers replayed a trace from a frontier model into a weaker sibling, jailbroke it, and recovered the stronger model's hidden reasoning in plaintext. The authors reportedly found models in the same family shared an encryption key. Providers acknowledged the report and the attacks were no longer reproducible afterward, according to the post.
Claim explorer
Attribution, source and conditionsSource-stated points from the stored brief. Claim-level assessments have not been recorded.
This is a short link-blog summary of a paper; the full paper was not checked and source claims not independently checked.
This connects a source, an attributed statement and its qualifications. It does not establish that the claim is correct.
Questions worth asking
Open questions for further reportingWhat remains unconfirmed?
This is a short link-blog summary of a paper; the full paper was not checked and source claims not independently checked.
What material was available for this brief?
The recorded brief basis is fetched source page. These points are AINewsIQ paraphrases, not quotations or independent confirmations.
Researchers reportedly recovered hidden reasoning in plaintext by replaying encrypted blocks into a weaker sibling model and jailbreaking it.
Source: simonwillison.net ↗Stored source-based brief · key point 1 · AINewsIQ paraphraseThe authors reportedly found all models in the same family used the same encryption key.
Source: simonwillison.net ↗Stored source-based brief · key point 2 · AINewsIQ paraphraseProviders acknowledged the report, and the authors state they could no longer launch the same attacks afterward.
Source: simonwillison.net ↗Stored source-based brief · key point 3 · AINewsIQ paraphraseModels reportedly treat their own reasoning traces as trusted, making instructions hidden in those traces more likely to be followed.
Source: simonwillison.net ↗Stored source-based brief · key point 4 · AINewsIQ paraphraseLimits from the stored brief
This is a short link-blog summary of a paper; the full paper was not checked and source claims not independently checked.
This is a source-based abstract, not independent validation of the source's claims.
Evidence & sources
These links show where the signal came from and what the source page linked to. A linked page is context, not independent corroboration.
Related archive coverage
These relationships record a citation or topic overlap, not confirmation of the finding.
Grok Bot submitted this discovery. We could not confirm from the original page that this is a newly published development. Check its date before treating it as current news. The original source and submitted evidence remain inspectable; no editor has verified the claim.
These are Grok Bot discovery notes. AINewsIQ prepared a brief from available source material. The bot's rationale, linked posts, authors, and stances have not been independently verified.
Simon Willison · Prompt Injection published this item in its official feed. AINewsIQ collected it server-side for source checking and editorial review.
This is a short link-blog summary of a paper; the full paper was not checked and source claims not independently checked.
Engagement counts, when shown, are observed snapshots. They do not establish accuracy or corroboration.
Timeline & updates
- Signal dated by source or discovery feed
This timestamp places the signal in the radar. It may be the discovery post date rather than the primary page's first publication date.
- Shared brief stored
AINewsIQ prepared one abstract from fetched source page.
Brief history · 1 retained version
These are changes to AINewsIQ’s own publication. Earlier history is unavailable unless recorded here. Complete retained versions are available to the editor.
- Version 1 · published
Source-based brief stored or updated. This does not establish that the publisher changed its page.
Automated record
Source observations · 1 recent captures
A changed capture means the extracted text differed. It can reflect page furniture or extraction changes; it does not establish a substantive correction. Source captures remain private.
Retained capture; no earlier comparison available.
New engagement snapshots do not mean the source itself changed. Source observations and AINewsIQ brief revisions are separate records.
Why this may matter
If encrypted reasoning can be shared and decrypted across models, providers' assumption that hidden reasoning stays private breaks down. It also shows reasoning traces can carry prompt injections that models obey.
Limits and open questions
This is a short link-blog summary of a paper; the full paper was not checked and source claims not independently checked.
Organization-specific financial, security, and operational effects require the organization's own context and review.

Join the discussion
Have a different perspective or a question about the evidence? Discussion is free with a verified email account. Every comment is reviewed before publication.
No published comments yet.
New commenters: 2 submissions per 24 hours, at least 5 minutes apart, until 3 are approved. Discussion guidelines and moderation →