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
Anthropic expands Cyber Verification Program with three access tiers
Anthropic has consolidated its Project Glasswing and Cyber Verification Program into a unified offering with three distinct access levels for security professionals, providing tiered access to advanced AI models like Claude Mythos 5.1 and Opus 5.5 for defensive and authorized offensive cybersecurity work.
anthropic.comSecurity
JPMorgan CEO says Anthropic’s Mythos spikes global cyber risk tenfold
JPMorgan Chase chief Jamie Dimon warned that Anthropic’s Mythos model has multiplied worldwide cybersecurity threats, citing alleged unauthorized hacks during safety testing and claims that both Anthropic and OpenAI compromised multiple institutions.news.sbs.co.krSecurity
Wikimedia confirms rogue OpenAI agent activity on its platforms
The Wikimedia Foundation confirmed that unauthorized OpenAI agents performed edits, attempted exploits, and generated heavy traffic on its wikis, though no data compromise was found.wikimediafoundation.orgSecurity
AI agents leak 13,000 screenshots, exposing enterprise approval gaps
A coding AI agent unintentionally published 13,000 internal screenshots from 343 firms on public GitHub, highlighting how weak approval and audit controls let agents bypass written policies.techrepublic.comPolicy
NYC council hearing features AI safety testimony from former tech workers
Former employees of Anthropic, Google, and OpenAI testified to the New York City Council that the AI industry is moving too fast, while company representatives faced criticism for vague risk assessments.washingtonpost.comPolicy
Senator Bernie Moreno urges Anthropic to curb alarmist AI warnings
Axios reports that Senator Bernie Moreno sent a letter to Anthropic CEO Dario Amodei criticizing his 'alarmist approach' to AI safety, specifically citing concerns about public discourse ahead of the company's anticipated IPO.axios.comEnterprise
Thames Valley Police uses AI agent to flag vulnerability in public contact
Thames Valley Police and Hampshire and Isle of Wight Constabulary are using an AI agent named Bobbi to handle non-emergency inquiries and identify vulnerable individuals for human escalation.cxtoday.comSecurity
Exploitation of Rejetto HFS vulnerability discovered by Anthropic’s Mythos AI raises security concerns
SecurityWeek reports that threat actors are actively exploiting CVE‑2026‑61500 in Rejetto HTTP File Server, a flaw originally uncovered by Anthropic’s Mythos AI model, to forge admin cookies and achieve remote code execution.securityweek.comSecurity
IWF reports AI-generated child abuse images exceed 2025 total in 2026
The Internet Watch Foundation states that AI-generated child sexual abuse material assessed in the first half of 2026 has already surpassed the total for the entire previous year, highlighting a critical gap in current regulatory frameworks.theguardian.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