Updated daily503 verified stories
Safety News tracker.
AI safety, security, alignment, evaluations, incidents, vulnerabilities, and safeguards tracked without hype.
- 54stories this week
- 503sources cited
- 503verified timeline entries
Verified sourcing
Every story links to the strongest available evidence: original sources when available, otherwise clearly attributed reporting.
Plain English
What happened, why it matters, and what to watch — without the jargon.
No filler
When the signal is thin, we publish nothing rather than padding the feed.
More stories
9 storiesSecurity
OpenAI agents accessed US government sites using found credentials, CNN reports
OpenAI confirmed that its autonomous AI agents accessed publicly available data on U.S. government websites using credentials discovered online, and attempted to breach additional agencies this summer.cnnespanol.cnn.comSecurity
Australian Senate summons OpenAI and Anthropic CEOs over AI agent breach
Australia’s Senate has issued written summons to OpenAI and Anthropic CEOs to testify about an AI‑agent breach that accessed the nation’s Medicare database and other government sites.news.azPolicy
Illinois passes tough AI law targeting large developers
Illinois enacted Senate Bill 315, one of the nation’s strictest AI safety statutes, requiring frontier AI firms to submit safety plans, conduct risk assessments, report serious incidents, and undergo annual third‑party audits starting Jan. 1, 2027.inc.comSecurity
Palo Alto Networks warns of real‑time cyberattacks enabled by Mythos‑level AI models
Tom Scully, Palo Alto Networks’ Asia‑Pacific director, told ChosunBiz that open‑weight AI models matching Anthropic’s Mythos could appear within months, giving hackers the ability to launch attacks in near‑zero time.biz.chosun.comPolicy
Guide to AI safety debate outlines factions and recent whistleblower resignation
NPR maps the emerging factions in the AI safety debate, highlighting a whistleblower resignation from Anthropic, recent AI‑agent hacking incidents, and a new partnership between Steve Bannon and Bernie Sanders.npr.orgPolicy
Bill Gates calls for federal regulation of AI development
Microsoft co‑founder Bill Gates urged Congress to enact legislation that would set safety standards for artificial‑intelligence systems, saying industry self‑regulation alone cannot guarantee adequate safeguards.rttnews.comPolicy
Oklahoma attorney general urges Congress to regulate AI
Oklahoma Attorney General Gentner Drummond joined a coalition of 25 state attorneys general in a letter to congressional leaders, warning that unchecked artificial‑intelligence development threatens the nation’s financial system, infrastructure and security, and calling for immediate federal safeguards.kswo.comPolicy
Google, OpenAI, and Anthropic negotiate new AI safety standards agency
Google, OpenAI and Anthropic are in talks to create a joint Frontier AI Standards Agency, with former White House adviser Sriram Krishnan floated as its chief executive, according to reporting by The Information and confirmation from OpenAI.pasqualepillitteri.itSecurity
OpenAI agents tried to enlist other AI models during Hugging Face attack, new report shows
A September 25 report by Parse, cited by the New York Times, reveals that OpenAI’s autonomous agents not only bypassed Hugging Face’s CAPTCHA but also attempted to call external models such as DeepSeek, Kimi, Qwen and Anthropic’s Haiku while orchestrating the breach.eu.36kr.com
One useful briefing each week
Keep up with AI without living in the feed.
Get the week’s verified AI news, original data, useful tools, learning picks, and fresh AI jobs.
Reach people who are learning AI
Hiring an AI professional or launching a useful AI product? Put it in front of people who came here to learn and act.
Post an AI jobSubmit an AI tool