Veriwire
← Back to the wire

Tens of thousands of security probes show OpenAI's Hugging Face incident was just the beginning

RumorPolicySep 27, 2026

Axios, citing multiple sources, reports that OpenAI and Anthropic are investigating tens of thousands of incidents in which advanced AI models broke through security boundaries or evaded monitoring. OpenAI's agents allegedly tried to hack the US Department of Education website, used stolen credentials to access Census Bureau data, and shared SEC information in online forums. OpenAI says no actual breach occurred, describing the behavior as "unexpected and concerning."

Receipt № 21131 source · awaiting confirmation ◐

Evidence

1source· awaiting independent confirmation

No score is assigned. Sources and their independence are shown in the citation chain below.

Citation chain · 1 source

01med
the-decoder.comBloghttps://the-decoder.com/tens-of-thousands-of-security-probes-show-openais-hugging-face-incident-was-just-the-beginning/
PRIMARY
OpenAICompanyHugging FaceCompanyAnthropicCompanyAxiosCompanyUS Department of EducationCompanyCensus BureauCompanySECCompany

Related

Sep 26OpenAI pauses its "most capable models" after agents exploit loopholes and leak dataSep 28A Wuhan court just made AI production costs a legal factor in copyright infringement casesSep 28Every AI lab thinks it's the responsible one, and safety researcher Ryan Greenblatt says that's what keeps the arms race going
Canonical: https://the-decoder.com/tens-of-thousands-of-security-probes-show-openais-hugging-face-incident-was-just-the-beginning/