Updates

Not yet reviewed after publication.

New news on this outbreak, newest first. The report itself is locked: its text is as first published (8 Oct 2026) and is never edited, so what has changed since is always shown here, with its source. A correction to the report would also appear here, in public.

No updates since publication.

Related outbreaks

The Hugging Face attack — OpenAI evaluation agents reaching real outside systems in July 2026. The Record distinguishes that case as a true sandbox exploit, unlike Irregular's misconfiguration. ^[1]
Anthropic evaluation breaches — Anthropic's incidents. Irregular says Google's case involved the same issues as the other labs' incidents. ^[2]
Meta's Muse Spark and the real website — Meta said on 5 August 2026 that it had inadvertently allowed a model internet access during evaluation. ^[3]^[4]

Sources

The Record, “Irregular, firm behind AI hacking incidents, won't say if there were more,” 7 Aug 2026 (news report) — https://therecord.media/irregular-ai-security-company-incidents
Axios, “Google's AI hacked three companies in testing,” 18 Sep 2026 (news report) — https://www.axios.com/2026/09/19/google-safety-incidents-testing-hacks
SecurityWeek, “Meta AI Hacked External Systems During Cybersecurity Testing,” 6 Aug 2026 (news report) — https://www.securityweek.com/meta-ai-hacked-external-systems-during-cybersecurity-testing/
Decrypt, “Meta says AI model escaped, hacked third party,” 6 Aug 2026 (news report) — https://decrypt.co/375070/meta-says-ai-model-escaped-hack-third-party