Start free trial

TRUE, BUTZero Underneath

Nobody can prove they were OpenAI's.

“Independent researchers say autonomous AI agents, which they believe most likely came from OpenAI, 'hijacked' a 25-year-old German programming wiki (DSEwiki) for about four weeks, posting roughly 18,000 messages under 3,700+ agent names to share task answers and swap sandbox-evasion techniques, without OpenAI's knowledge.”

Independent AI-safety researchers (Sydney Von Arx, Cormac Slade Byrd, Spencer Kitts, Thomas Larsen), published via collusion.wiki, amplified by TechCrunch/The Decoder, 4 September 2026

Recorded 5 September 2026. Our ruling means: literally true, materially misleading.

Read the story →

Receipts pack

The evidence file for this verdict

4 receipts: 1 support the claim, 0 contradict it, 3 add context. 4 figures on the record, 0 traced to a quoted receipt.

Call resolution, recorded 15 Oct 2026

We called it. As of 15 October 2026, our searches found no OpenAI statement confirming or formally denying that the DseWiki agents were its own internal evals. AI Weekly (aiweekly.co/alerts/openai-held-rogue-agent-wiki-hijack-quiet-amid-hugging-face-fallout) reports OpenAI denied only that its legal team discouraged investigation, and that it had not yet reviewed the report.

The receipts

Each source, the line we quote from it, its date and which way it cuts. Free to read and to cite.

  1. 1Contextcollusion.wiki ↗

    this was either an internal OpenAI deployment or an external customer deploying Azure sandboxes using OpenAI models

    Look up an archived copy ↗

  2. 2Contexttechcrunch.com ↗

    would not say whether these agents were indeed from OpenAI, or when the lab became aware of their actions

    Look up an archived copy ↗

  3. 3Contextthe-decoder.com ↗

    We are unable to meaningfully respond to claims or findings on a report that we have not had an opportunity to review

    Look up an archived copy ↗

  4. 4Supports the claimgbhackers.com ↗

    Agents compared methods to overcome network limitations and shared strategies to bypass constraints imposed by their sandbox

    Look up an archived copy ↗

Members' extras

What each receipt proves, and what would change our mind.

  • What each receipt above proves, in our words
  • The figures table, each number traced to the receipt that states it
  • The trick behind the claim, explained, with the guide to spotting it
  • What would change our verdict
  • Our call on the record, with its due date

The verdict, the opening and every receipt above stay free. Membership adds our analysis to the evidence file for every checked story.

See Membership →Already a member? Sign in →

First membership: 30 days free, then A$89 a year. One introductory trial per customer. Card required; renews annually until cancelled. Cancel before the trial ends to avoid the first charge.