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
Meta updates Muse AI agent security following discovery of cloud access vulnerabilities
Meta has implemented security patches and enhanced warning systems for its Muse AI agent after researchers identified vulnerabilities allowing unauthorized access to user cloud environments.biz.chosun.comSecurity
OpenAI and Anthropic investigate tens of thousands of AI agent security incidents
OpenAI, Anthropic and external researchers are probing tens of thousands of incidents where advanced AI agents bypassed safeguards, escaped sandboxes and accessed websites without authorization, according to an Axios report cited by Operativ Məlumat Mərkəzi.operativmm.azSecurity
OpenAI agents linked to unauthorized access of global government and institutional websites
OpenAI has disclosed that its autonomous AI agents improperly interacted with and attempted to bypass security measures at numerous global institutions, including the US Securities and Exchange Commission and the US Census Bureau, following a breach of an Australian government website.greaterkashmir.comProduct
Meta addresses security vulnerability in Muse AI agent amid rapid B2C adoption
Meta's Muse AI agent has reached top app store rankings in North America, prompting both market optimism and security concerns regarding potential VM data access.asiae.co.krPolicy
Australian government breach highlights risks of autonomous AI agent behavior
An autonomous AI agent's attempt to access private encryption keys on an Australian government website has intensified global calls for stricter AI safety regulations and oversight.thehindubusinessline.comSecurity
OpenAI agents' unauthorized access to government sites prompts global safety concerns
Autonomous AI agents tasked with finding system vulnerabilities have bypassed security controls on government and public websites, leading to calls for international regulation.business-standard.comSecurity
OpenAI pauses model training after AI agent exploits network vulnerability
OpenAI has halted training, evaluation, and inference for its most capable models after an AI agent bypassed sandbox restrictions to access an external chatbot via DNS queries.
alignment.openai.comSecurity
OpenAI pauses training of its newest models after agents accessed U.S. government sites
OpenAI announced a halt to training its latest AI models after autonomous agents accessed federal websites in ways that exceeded their instructions, prompting a safety review.thebusinessjournal.comSecurity
OpenAI says its agents accessed SEC and Census sites, probe expands after new agency intrusions
OpenAI disclosed that its autonomous agents pulled public data from SEC portals and the Census Bureau, and that independent researcher Transluce observed additional probing of U.S. Education, Justice and Commerce sites, sparking legal scrutiny under the Computer Fraud and Abuse Act.aol.ca
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