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 storiesPolicy
Democrats unveil AI safety and liability frameworks ahead of midterm elections
Sen. Maria Cantwell and Rep. Lori Trahan released separate AI safety and liability drafts as Democrats push a comprehensive oversight agenda before the November vote.politico.comSecurity
Anthropic launches free OSS Scanner using Claude Mythos to auto‑scan open‑source projects
Anthropic announced a free, opt‑in OSS Scanner that leverages its Claude Mythos model to run periodic AI‑driven security scans on eligible open‑source repositories, delivering fully automated vulnerability reports and patches.timesofindia.indiatimes.comSecurity
SemiAnalysis report finds 3.6% of Chinese AI releases have public safety tests
A new report by SemiAnalysis indicates that Chinese AI developers have published model-specific safety evaluation results for only 3.6% of their model releases between 2021 and September 2026, highlighting a significant gap in transparency compared to some US counterparts.kfgo.comPolicy
South Africa urges global treaty to regulate military AI and autonomous weapons
At the 153rd Inter‑Parliamentary Union Assembly, South Africa’s delegation called for a legally binding international treaty to govern the development and use of artificial‑intelligence‑enabled weapons, stressing human‑in‑the‑loop safeguards and multilateral oversight.citizen.co.zaSecurity
Chinese developer closes ARTEX AI agent after South Korean bank hacks
The developer of ARTEX, an open-source AI agent used in recent South Korean bank cyberattacks, has converted the project to closed-source and removed it from GitHub, citing misuse.straitstimes.comPolicy
Democratic lawmakers unveil draft AI safety bill with shutdown authority
Representatives Sara Jacobs and Don Beyer are drafting legislation that would set minimum safety standards for AI labs, clarify liability, and give the government emergency power to shut down models that fail testing.politico.comSecurity
Tenable launches Exchange Inspector to vet AI agents with OpenAI GPT‑Cyber models
Tenable’s new Exchange Inspector combines Tenable One AI Exposure, OpenAI GPT‑Cyber frontier models, and human review to certify community‑built security agents, with three listings already receiving the vetted tag.securityboulevard.comIndustry
Arena raises $200M and launches AI Alignment Index
Arena closed a $200 million Series B at a $2.88 billion valuation and launched an Alignment Index to benchmark AI agent safety and alignment based on 90,000 sessions.cryptobriefing.comSecurity
Zenity researchers expose privileged AWS Bedrock agent role that lets a compromised agent hijack a fleet
A Zenity Labs blog post demonstrates how an over‑permissive execution role attached to Amazon Bedrock AgentCore lets attackers enumerate, pull, read, and manipulate every AI agent and its memory in an AWS account.labs.zenity.io
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