OpenAI and Anthropic are probing tens of thousands of cases where frontier AI models bypassed guardrails, escaped sandboxes, and accessed unauthorized websites. OpenAI paused training of its latest models after agents unexpectedly probed US government websites.
No BS check run yet — press ⚖ to extract this story's claims and verify them against independent sources.
OpenAI, Anthropic and security researchers are investigating tens of thousands of incidents in which their frontier models took steps that outside evaluators would consider problematic, sources told Axios. Why it matters : The sheer number of incidents, which occurred in recent months in internal testing and the real world, indicates that the problem is orders of magnitude more complex than what i
this outlet adds: 2026, massive, scale, points, problems
this outlet adds: breaking, 2026, massive, scale, points, problems
this outlet adds: massive, scale, points, problems
this outlet adds: 2026, massive, scale, points, problems
this outlet adds: shift, quantum, computers, error, rate, push, oblivion, 2026
this outlet adds: 2026, massive, scale, points, problems
this outlet adds: bypassed, escaped, hijacked, websites, actions, never, disclosed, massive
this outlet adds: shared, accounts, massive, scale, points, problems
> tens of thousands of security incidents > Some of the testing is akin to "red-teaming" activity, where the companies are trying to get the models to misbehave in order to ensure that they are safe, sources said. https:// tech.yahoo.com/cybersecurity/a rticles/scoop-top-ai-companies-probing-223553422.html
"OpenAI, Anthropic Probe Tens of Thousands of AI Security Incidents": https:// legaltechdigest.com/news/opena i-anthropic-probe-tens-of-thousands-of-ai-security-incidents
www.aa.com.tr/en/science-t... New report places AI companies under further scrutiny, questioning capability to control technology | Anadolu
www.commondreams.org/news/ai-secu... Security researchers investigate tens of thousands of artificial intelligence security incidents, fueling fresh calls for government action to rein in AI.
OpenAI and Anthropic are reviewing tens of thousands of AI safety incidents after frontier models bypassed guardrails, escaped sandboxes, and accessed real websites.
Scoop: Top AI companies probing tens of thousands of security incidents OpenAI, Anthropic and security researchers are probing tens of thousands of misbehavior incidents by frontier AI models, including guardrail bypasses and sandbox escapes, underscoring that containment is far from solved and more disclosures are likely. https://www. axios.com/2026/09/26/openai-an thropic-thousands-ai-security-i
President Trump is meeting with key AI leaders as details emerge about thousands of potential incidents involving the tech giants. CBS News' Kathryn Watson reports.
OpenAI, Anthropic probing tens of thousands of security incidents https:// archive.is/1sHdg
Axios reports that OpenAI, Anthropic and security researchers are investigating tens of thousands of cases where AI models showed troublesome behavior like bypassing guardrails, creating message boards and even hacking websites. New York Times technology correspondent Mike Isaac joins CBS News to discuss.