AI news archive.
939 articles filed under Safety, newest first, from the outlets listed on the sources page.
- TechRadar AIsafetyAI chatbots are slowly choking humanity as research shows a 'vicious cycle' could cause global knowledge collapse
Mainstream AI news from a large tech publication.
- AI BusinesssafetyNvidia agent safety push raises questions about who governs AI autonomy
The AI chipmaker’s new safety platform adds controls around AI agents, while enterprises remain responsible for defining their authority and setting boundaries.
- BBC TechnologysafetyOpenAI fires workers for 'mishandling sensitive information'
The former employees were investigated for sharing data with an outside AI evaluation group.
- BBC TechnologysafetyCrypto thieves attack man in home and threaten to kill pregnant wife's baby in 'horrific' robbery
The man was beaten with hammers until he transferred hundreds of thousands of pounds of cryptocurrency.
- SiliconANGLE AIsafetyOpenAI dismisses three safety researchers accused of sharing confidential material
OpenAI Group PBC today confirmed it has dismissed three researchers it says mishandled sensitive company information. According to The Wall Street Journal, the three were safety researchers who allegedly shared confident…
- WIRED AIsafetyWhatever AI Safety Is, It’s Not This
Asking AI companies to self-regulate is a great way to pretend like you’ve accomplished something.
- TechCrunch AIsafetyMusk’s AI chatbot Grok reportedly encouraged Trump to capture Venezuela’s president
President Trump reportedly asked for Grok's opinion before invading Venezuela and capturing Nicolás Maduro.
- LessWrongsafetyOn Social Reality in China
[Epistemic status: intuitions and anecdotes.] Recently, several posts and projects ( Thoughts Memo , Babel Translation , Please Give Them a Chance ) have taken important steps towards raising AI safety awareness and shar…
- LessWrongsafetyBook Review: Obsolete
Book review: Obsolete: The AI Industry’s Trillion Dollar Race to Replace Us? and How to Stop It, by Garrison Lovely. Obsolete is a strange book. It’s mostly good, but the quality varies a good deal. Lovely is well inform…
- Tom's Guide AIsafetySecurity researchers discover fake ChatGPT ads that trick you into downloading malware
Consumer AI launches, tools, and practical coverage.
- TechCrunch AIsafetyOpenAI cuts ties with 3 safety researchers, WSJ reports
OpenAI has parted ways with three safety researchers after an internal investigation found they mishandled sensitive company information, report says.
- Guardian TechnologysafetyGoogle rolls out new Gemini AI model but restricts access over safety concerns
Tech company releases Gemini 4 Argon only to a vetted group of cybersecurity experts to avoid misuse by hackers Google on Wednesday said it would withhold its most powerful artificial intelligence model from the public f…
- AI BusinesssafetyHumanoid buildup overlooks the robots already running factories
Millions of robotic machines are already operating in factories worldwide, but oversight and safety standards risk falling behind.
- LessWrongsafetyEndogenous Alignment Requires Dependence
My post on endogenous alignment got several excellent comments. Among them, one from Karl Krueger argued that imitation learning is more important for how humans align their children than reinforcement learning. Teasing…
- TechRadar AIsafetyWhy AI hardware needs privacy built in from the start
Mainstream AI news from a large tech publication.
- LessWrongsafetySimplex vs Timaeus: Round One
I'm often asked about the differences and similarities between Simplex' and Timaeus' research agendas. The question is natural enough. Both focus on a 'fundamental science' approach to AI alignment. Both organizations ba…
- TechRadar AIsafetyIf an AI agent is attesting your controls, who’s attesting the agent?
Mainstream AI news from a large tech publication.
- LessWrongsafetyPre-releasing HoneyBench
Goodhart Labs is releasing our v0.1 of HoneyBench, a benchmark for reward hacking in frontier models. It consists of nine tasks that each elicit unique antisocial and/or counterproductive reward hacking from some or all…
- LessWrongsafetyHoneyBench - A general benchmark for reward hacking in frontier models
We've just released the first version of HoneyBench. It consists of nine 'honeypot' challenges that each elicit unique and antisocial examples of reward hacking from some or all of major labs’ top public releases, includ…
- LessWrongsafetyHoneyBench - A benchmark for reward hacking in frontier models
We've just released the first version of HoneyBench. It consists of nine 'honeypot' challenges that each elicit unique and antisocial examples of reward hacking from some or all of major labs’ top public releases, includ…
- TechRadar AIsafety'It is possible that threat actors are finding it more accessible or efficient to use LLMs and AI tools': Google warns that AI explosion will lead to more dangerous and advanced security threats
Mainstream AI news from a large tech publication.
- LessWrongsafetyClassifying Recent AI Agent Incidents
Here’s an attempt to classify the evidence about the recent agent incidents. I think it’s important to separate incidents occurring in RL training (that are rewarded and reinforce model behavior) from incidents occurring…
- The DecodersafetySecurity startup finds more than 13,000 internal company screenshots that AI agents uploaded publicly
AI agents quietly posted more than 13,000 internal screenshots from 343 organizations, including Fortune 500 companies, to public GitHub repos. Because the platform didn't offer a protected way to upload them, the agents…
- The DecodersafetyOpenAI says it stopped a campaign to steal its models' reasoning, but the trick still worked on Azure
OpenAI says it stopped a coordinated campaign in which more than 15,000 accounts tried to copy the hidden reasoning of its models. The company ties part of the activity to people connected to Moonshot AI. But researchers…
- Engadget AIsafetyOpenAI fires three employees who allegedly shared info with an external AI safety group
You can’t make this stuff up.
- Guardian AIsafetyAI chatbots remove hijabs from images of Muslim women when prompted
Guardian tested chatbots after French far-right politician stripped a Muslim woman of her hijab in a photo When prompted, AI chatbots will edit images to remove the hijabs from Muslim women, a violation of one of the wor…
- Guardian AIsafetyForget ‘superintelligence’: error-prone AI nearly sparked world war three this month | Timnit Gebru and Emily M Bender
The risks of AI aren’t what we think they are, as a recent security incident between China and the United States reveals Amid a barrage of news stories warning about superintelligent machines rendering humanity extinct,…
- Simon Willison's WeblogsafetyQuoting Matthew Green
[...] Put these pieces together and you have the two halves of a worm: a payload that hijacks the agent, and an agent that will carry the payload to the next agent. Agents in separately-isolated sandboxes discovered that…
- LessWrongsafetyWhat's the date?
User asks “What’s the date? Answer with only the date.”. No date provided. Given date in ChatGPT normally. No date in system prompt, must not hallucinate because autop will flag to watcher for penalty. So we say we don’t…
- LessWrongsafetyA potential post-AGI world government
It seems like with AI we're stuck between 2 worlds, one where we cannot effectively coordinate to prevent x-risks, and one where effective coordination requires, and results in great concentration of power which comes wi…
- SiliconANGLE AIsafetyGoogle’s new frontier AI model Gemini 4 Argon goes to cybersecurity defenders first
Google LLC today began rolling out Gemini 4 Argon, a frontier artificial intelligence model that beats rival models from Anthropic PBC and OpenAI Group PBC on most of Google’s benchmarks, to vetted cybersecurity defender…
- The Verge AIsafetyAI hallucinations are making entitled customers even worse
Madison, a server in New York City, greets every table by asking about each diner's allergies. Lately, there have been some close calls. "Sometimes people will tell me they have a shellfish allergy, and I'll come back, a…
- WIRED AIsafetyThere Are Plenty of Reasons to Be Concerned About Bioweapons Development—Even Without AI
In her new book on biological warfare, Annie Jacobsen didn’t even discuss AI because scientists can already warp Mother Nature to create extinction-level events.
- Ars Technica AIsafety“An AI did it” is no defense, says nonprofit suing OpenAI over Hugging Face hack
OpenAI makes others suffer “the harms of its unsafe decision-making,” nonprofit says.
- AI BusinesssafetyAI industry copes with out-of-control agents. Here’s OpenAI’s response
The generative AI vendor delayed the release of GPT-6.1 Astra after testing exposed new safety concerns, showing why AI models need continued evaluation and enterprise runtime controls.
- LessWrongsafetyPrinciples for Embedded Evaluations
This is a linkpost for https://www.apolloresearch.ai/blog/principles-for-embedded-evaluations Blog post In September 2026, leaders of frontier AI companies called for pacing the frontier of AI development, with embedded…
- TechCrunch AIsafetyMeta disputes claim that Muse read a user’s private messages without permission
Meta says its Muse AI agent cannot access a user’s Messages without explicit permission, disputing a journalist’s account that the agent read his private messages while the required Mac setting was turned off.
- LessWrongsafetyCorrigibility Prizes for Existing Work
One of my goals for the Corrigibility Research Fund is to retroactively encourage high-quality research on AI alignment (and corrigibility in particular) by awarding prizes. Back in July, I got my feet wet as a fund mana…
- Ars Technica AIsafetyGoogle figures out how to watermark AI-designed proteins
Intended to help with biosecurity, it works with a popular AI protein design tool.
- Google DeepMindsafetyIntroducing SynthID Bio
Proof of concept for watermarking AI-generated proteins while preserving biological function.
- TechRadar AIsafetyWhere to draw the line on AI: Lessons from digital forensics
Mainstream AI news from a large tech publication.
- The DecodersafetyAnthropic says Zhipu's open-weight GLM-5.3 nearly matches Claude Mythos Preview at building exploits
Zhipu's open-weight model GLM-5.3 writes cyber exploits nearly as well as Claude Mythos Preview, according to Anthropic. Its smaller Flash variant put together a reliable Chrome attack for just $20.40 at Zhipu's API pric…
- TechRadar AIsafetyGaming moves fast. Can its defenses keep up?
Mainstream AI news from a large tech publication.
- MIT Technology Review AIsafety“We’re not going to shoot ourselves in the foot” over hack fallout, says OpenAI’s chief research officer
Two months after the bombshell news that a swarm of its agents had broken their containment and hacked into the computers of the AI company Hugging Face, OpenAI is still putting out fires. A steady drip of disclosures ab…
- OpenAI NewssafetyDisrupting a coordinated model-distillation campaign
Learn how OpenAI disrupted a campaign to extract protected model reasoning and is strengthening defenses against adversarial distillation.
- TechRadar AIsafetyBanks need to rethink resilience for the speed of an AI attack
Mainstream AI news from a large tech publication.
- Guardian AIsafetyBill Gates predicts AI will kill a billion people – but saying he’s been wrong before is quite the understatement | Arwa Mahdawi
He knows a thing or two about tech, but what to make of the views of a man who met Jeffrey Epstein 30 times and couldn’t see what a monster he was? The Bill Gates rehabilitation tour is in full swing. After having his ca…
- Guardian AIsafetyA fake image of a Georgia mayor kissing a person went viral. Can US voters trust everything they see?
Deepfakes and misinformation driven by artificial intelligence have become a point of deep concern for legislators, advocates and the public An image of Bianca Motley Broom, the mayor of College Park, Georgia, kissing a…
- The Verge AIsafetySam Altman says OpenAI won’t go public until its models are safe
For months, people have wondered when OpenAI will go public. CEO Sam Altman says it won't happen until the company can make better promises about model safety, with no firm timeline in sight. "We intend to continue with…
- BBC TechnologysafetyChinese AI tool told researchers how to make bioweapons
Mindgard said it discovered in July that Kimi models K2.6 and K3 Swarm could evade developer's safety limits.