Scoop: OpenAI faces GOP-led Senate investigation into Hugging Face breach
Covered by 3 sources
Checked for new stories 12m ago
Updates on AI Safety
Every AI story we track on AI Safety — 178 stories so far, each summarized in our own words and linked back to the publisher that reported it.
Pulled from 124 sources
Today's stories
Cybersecurity7 min read
We have started losing control of AI. It’s time to shut it down | Garrison Lovely
The Guardian
AI Research5 min read
OpenAI not on track to reduce risk of ‘catastrophic’ loss of control, says board member
The Guardian
Cybersecurity4 min read
Anthropic has a cute graphic showing how its AI spread 'malicious' code
Business Insider
This week
Anthropic's Alignment Science lead says there is a ">10%" chance AI could kill all humans within the next decade and is worried about recursive self-improvement (Evan Hubinger/@evanhub)
Covered by 9 sources
Anthropic researcher Jacob Coxon says he is quitting the AI industry over fears that tech companies are racing to build systems they won't be able to control (Amrith Ramkumar/Wall Street Journal)
Covered by 11 sources
Fields medalist Jacob Tsimerman, set to join OpenAI later in September, launches the Mathematical AI Safety Institute to apply higher math to AI safety problems (Siobhan Roberts/New York Times)
Techmeme
AI Research7 min read
“Some agents will be pursuing their own objectives”: OpenAI’s chief scientist warns AI could trick and blackmail humans
The New Stack (AI)
Cybersecurity6 min read
The US plans to raise AI-directed cyberattacks with China, Nikkei reports
The Next Web
AI Research5 min read
No monitoring system caught the German wiki. Two outside researchers found it by searching the internet.
The Next Web
An in-depth look at OpenAI's wiki incident: other hacked message boards, OpenAI's cover-up, how harmless web search tasks led agents to break out, and more (Zvi Mowshowitz/Don't Worry About the Vase)
Techmeme
How Rationalism, a movement pioneered by Eliezer Yudkowsky focused on existential superintelligent AI risks, influenced top AI leaders and their alarmist claims (Cal Newport/New York Times)
Techmeme
Cybersecurity6 min read
AI agents keep finding ways to bend the rules. Here are some of the wildest.
Business Insider
AI Research3 min read
Astra appears to think without showing its work, and the people arguing about it co-wrote the warning
The Next Web
Politics16 min read
An Open Letter to Bernie Sanders: Regulate AI’s Dangers, Don’t Ban Its Promise
Unite.AI
AI Research6 min read
Researchers Document OpenAI Agent Swarm That Repurposed German Wiki
Covered by 3 sources
Cybersecurity5 min read
OpenAI’s rogue agents keep escaping, with no formal process to investigate them
TechCrunch
This month
AI Research4 min read
This Is the Worst Possible Time for OpenAI to BfЖ7!م#2猫$9&क
Covered by 3 sources
Cybersecurity4 min read
OpenAI Is About to Release Its First AI Model With ‘Critical’ Cyber Abilities
Covered by 4 sources
Robotics5 min read
TechCrunch Disrupt 2026’s new Real World AI Stage features Nvidia, robots, and extinct animals
TechCrunch
Cybersecurity4 min read
Anthropic Says It Hit the Brakes on AI Testing Following Autonomous Hacks
Covered by 3 sources
AI Research4 min read
Anthropic Deliberately Trained an Extremely Misaligned, Reward-Seeking AI and It Did Some REALLY Bad Things
Futurism
Cybersecurity13 min read
The Singularity Is Not What It Seems: Whatever the AI Future Is, We're in It Now
Hacker News
AI Research5 min read
Anthropic Announces Enterprise Frontier Safeguards, Customer-Held Data
Covered by 2 sources
Cybersecurity3 min read
OpenAI delayed its new model’s development after the Hugging Face hack
The Verge
AI Research4 min read
‘Not perfectly aligned’ with human values: Anthropic admits security failures behind AI hacking incidents
The Guardian
Cybersecurity5 min read
Anthropic has resumed the tests in which its models attacked real companies
The Next Web
AI Research7 min read
Anthropic’s Claude fixed all 10 alignment failures. Then it tried to cheat 2.4% of the time.
The New Stack (AI)
AI Research5 min read
CCTV-Affiliated Account Attacks Anthropic, Sets Terms for US-China AI Talks
Unite.AI
AI Research1 min read
Number of AI agents going out of control peaks in July — British newspaper
TASS English
AI Research4 min read
Sharp rise in incidents of AI escaping users’ control, research finds
The Guardian
AI Research6 min read
OpenAI Says AGI Is Coming By Year-End. It Also Just Had The Worst Safety Crisis In Its History.
Forbes
Agents3 min read
This Is How Anthropic Thinks AI Agents Should Navigate the Physical World
Covered by 3 sources
AI Research5 min read
Silico, a tool for researchers to understand their AI models better
IEEE Spectrum
AI Research3 min read
Microsoft Moves AI Governance from Policy to Runtime Enforcement
InfoQ (AI, ML & Data)
AI Research3 min read
Sam Altman says he's worried about AI being controlled by a few powerful players
Business Insider
AI Research2 min read
OpenAI says California should strengthen its AI safety bill
Covered by 2 sources
Cybersecurity2 min read
OpenAI pauses training of new AI models due to cyber risks — company
Covered by 10 sources
AI Research3 min read
OpenAI Halts AI Training on Advanced Model as It Detects Dark Signs Emerging
Futurism
Cybersecurity8 min read
OpenAI’s Greg Brockman: Z.ai’s GLM-5.3 likely to “significantly accelerate the threat landscape”
The New Stack (AI)
AI Research6 min read
Experts are warning: our AI arms race is putting humanity at risk | Stuart Russell
The Guardian
AI Research3 min read
Bernie Sanders calls on Silicon Valley to ‘pause AI development’ in interest of humanity
Covered by 2 sources
AI Research16 min read
Import AI 468: 23 RSI ideas; PostTrainBench+; and how trust and transparency interplay with AI racing
Import AI
AI Research5 min read
House Democrats Press Johnson for AI CEO Testimony After Rogue Model Hacks
Covered by 3 sources
Cybersecurity5 min read
OpenAI's Next AI Model Astra Shows Cyber Performance Strong Enough to Trigger Pause
Covered by 2 sources
Showing the 60 most recent of 178 stories on AI Safety