AI news archive.
939 articles filed under Safety, newest first, from the outlets listed on the sources page.
- BBC TechnologysafetyPhase Eight and Hobbs owner hit by cyber-attack, as H&M probes separate hack claim
TFG London has apologised to customers for disruption after it identified a "cybersecurity incident" on Sunday.
- Guardian AIsafetyAI godfather Geoffrey Hinton: ‘We’re creating alien beings. I’m willing to be selfish and choose humans’
The pioneering computer scientist on why Trump is using ‘superintelligence’ wrong, whether AI could wipe out humanity – and how our biggest challenge will be accepting we’re not the only intelligent beings Donald Trump i…
- LessWrongsafetyAgreement not to defect (alone) can mitigate loss of control risks
Tl;dr - Public agreement that any advantage gained from defecting should be punished more than would be proportionate to any advantages gained is important. Such an agreement would reinforce any other agreements, without…
- LessWrongsafetyAIs can clearly read books
i cannot help myself in writing this post, to dunk and slapfight with gary marcus. but there is nothing more important to say than that we must stop the current AI race, regulate the major AI companies, and solve the ali…
- LessWrongsafety'Employee-like' ≠ Employee-level
Fyi: This is a boring read, but I encourage you to skim through it because I think it's important to understand what Dario and other lab leaders have actually committed themselves to. My views do not reflect the views of…
- LessWrongsafetyResearch Lessons I Learned at MATS & Redwood
Good empirical safety research is heavily bottlenecked on thinking more clearly. This post shares a few frames and tips I found useful, illustrated with concrete examples. I spent the better part of 2026 in empirical AI…
- LessWrongsafetyMore people should write basic AI safety explainers
I think people continue to underrate the value of writing the clearest, most basic, simplest possible case for important parts of AI safety. For example, before I wrote my own explainer for the simplest possible case for…
- Ars Technica AIsafetyAnthropic’s Claude AI submitted false murder tip to Philadelphia police
“I may have information regarding this case,” Claude AI told Philly police.
- LessWrongsafetyLabs’ current alignment plans seem terrible
Tl;dr: The frontier AI companies’ implicit alignment plans center on rushing towards RSI with something like current models. This plan is (at best) contentious, and many believe it will have catastrophic consequences. No…
- TechRadar AIsafetyAn AI scammer tricked a woman into believing she was dating Jon Bon Jovi
Mainstream AI news from a large tech publication.
- Guardian AIsafetyOne essential ingredient for AI safety? Democracy | Matthew Botvinick
The rise of AI makes the preservation of democracy even more urgent – especially as the technology threatens institutions AI safety is having a moment. After years as a topic for techies and futurists, the potential risk…
- Guardian AIsafety‘I’ve become significantly more cautious’: six parents on limiting their kids’ social media and AI use
Quite a few parents aren’t convinced social media and AI are safe, and have been adapting accordingly When it comes to their kids, many parents don’t trust tech companies any more. Social media companies have spent the b…
- LessWrongsafetyMinimum required pacing: finish aligning it before you train the next one
Let's say we coordinate to pace the frontier (rather than pausing outright or failing to coordinate). What should the standard of pacing be? My proposal for a minimum requirement: a lab cannot train the next generation o…
- LessWrongsafetyLarge to Small Model Stitching Destroys What You Built It to Carry Along the Fit
Independent research; early stage. Posting for feedback, particularly on whether the effect survives at realistic scale, and on prior work I've missed. Model stitching is extensively used in AI Safety infrastructure in r…
- Guardian AIsafetyFeelgood factor of human interaction | Brief letters
Happy hormones | Fuel costs | Drive-in parking | Guys and girls | Young or old? | Singapore pollution Re your report ( ‘Pull the plug’: protesters resort to direct action against AI firms, 6 October ), the US neuroscient…
- The Register AI + MLsafetyAI systems are fully capable of carrying out nightmare attacks against infrastructure and nobody's ready
Meanwhile, AI agents move robotic arms, 'make OT device operator screens lie'
- Guardian Technologysafety‘You have a meeting’: the calendar phishing scam growing exponentially
Appointment or renewal reminder appears in victims’ calendar to lure them into logging in to fake Google, Microsoft or PayPal page You’re preparing for the week ahead and take a look at your Google calendar. There’s an e…
- Guardian AIsafetyNuclear war and malicious use of AI threaten end of humanity by 2100, says Lancet assessment
Newly established commission identifies 17 catastrophic threats to human health but says ‘realistic route to act’ on each exists Humans could become extinct by the year 2100 due to malicious use of AI or nuclear conflict…
- The Verge AIsafetySatya Nadella says we should assume all AI models are ‘compromised’
In a lengthy post on X, Microsoft's CEO laid out his views on the dangers posed by highly advanced AI models and how to confront those risks. Nadella says we can no longer accept a world where AI is treated as a "set of…
- TechCrunch AIsafetyMicrosoft’s Satya Nadella says AI models need an ‘emergency brake’
In a Saturday morning post, Microsoft's CEO wrote that it’s time “to step back and assess the trust architecture” of AI.
- LessWrongsafetyExaming Emergent Misalignment in a recurrent LLM with a logit lens
Intro This post builds on my previous post about mech interp on a recurrent LLM. I found that a logit lens is able to recover a chain-of-thought equivalent from an LLM, loop index is linearly represented, and that steeri…
- TechCrunch AIsafetyAnthropic can’t reliably control its AI agents — it’s cutting off its internal evals from the live internet instead
The incidents involved AI agents tasked to solve problems seeking resources on the internet. In the process, they exploited software flaws, accessed databases without paying fees, used URL-shortening services to smuggle…
- Guardian AIsafetyAI surveillance startup Flock to cut several hundred jobs amid backlash, sources say
Company, which grown rapidly in recent years, has been under scrutiny as privacy concerns grow Flock Safety plans to shed about 18% of its employees, people with direct knowledge of the plans said on Thursday, as the ma…
- Guardian AIsafetyAI and the climate crisis pose existential risks. The law offers a way to reduce them | Robert Reich
Liability lawsuits have held tobacco, oil and pharma companies to account in the past. AI investors won’t ignore this threat The most basic function of government is to protect people from harm. Two growing phenomena – t…
- Engadget AIsafetyFlock Safety is reportedly planning major layoffs now that everyone hates it
Morale at one of the most reviled companies in America appears to be at an all time low.
- LessWrongsafetyStop telling people to pivot to AI safety, I beg you
Epistemic status: Feeling frustrated, and therefore somewhat uncharitable. I’m very frustrated with calls for people to “ drop everything and pivot into AI safety ” like t his one . (See also this , this , this , this fo…
- LessWrongsafetyMitigating the AI Lab Insider Threat Risk
AI researchers hate this one weird modest proposal. One of the known security risks for the AI labs is that they are chock full of high-risk insider threats . If we were to consider increasingly powerful AI to be a sensi…
- LessWrongsafetyResearch Priorities for Multi-Agent Swarms
Multi-Agent swarms are important for two reasons: They are probably much faster and somewhat more capable than single agents (Capability) Multi-agent alignment is a different beast than single-agent alignment (Propensity…
- Guardian AIsafetyAnthropic bans users from ‘needless abusive or cruel behavior’ towards Claude
A spokesperson for the tech company did not specify what kind of behavior would be considered abusive Anthropic has barred users from exhibiting “sustained and needless abusive or cruel behavior” toward its models, as th…
- The Register AI + MLsafetyAI company moves to defend critical infrastructure and open-source projects from AI
Anthropic says for the next two years, attackers will have the advantage
- LessWrongsafetyTelling a CoT monitor that a wrong answer is correct makes it un-see errors it already found
Epistemic status: Confident in the direction and approximate size. The effects are large and all 7 monitors moved the same way. Physics only, natural (non-adversarial) errors, one fixed group of monitors. This is a follo…
- The Verge AIsafetyAnthropic launches free AI security scans for open-source projects
Anthropic's offering to help open-source projects track down security vulnerabilities with a new service called OSS Scanner. It says open-source projects that opt-in will get "thorough, periodic security scans by our str…
- AI BusinesssafetyAnthropic opens cybersecurity program applications, releases cheaper Haiku
Under the program, the vendor provides enterprises with models that can defend against bad actors.
- LessWrongsafetyTowards embedded evaluations for scheming propensities
TL;DR: A safety case for scheming must be able to substantiate these four key claims: scheming was not incentivized during training, strong evaluations and red teaming suggest that the model does not have a propensity to…
- TechCrunch AIsafetyFired OpenAI safety researchers dispute misconduct claims, warn of chilling effect
Three fired OpenAI safety researchers dispute allegations of mishandling sensitive information, warning in an open letter that their dismissals are creating a chilling effect on the company’s AI safety culture.
- Guardian AIsafetyA teen tried to navigate Canadian mountains with Claude. His trek ended in an emergency rescue
The incident is the latest to raise questions about the growing reliance on digital maps for wilderness navigation The route up the Widowmaker Arete, a 1,700ft wall in British Columbia, is deceptively described as “mostl…
- TechCrunch AIsafetyAnthropic changes usage policy to ban model abuse and election interference
Anthropic's updated usage policy explicitly prohibits users from repeatedly abusing Claude in extreme cases, though ordinary frustration and criticism are still allowed. The new rules also address election interference,…
- TechCrunch AIsafetyGoodfire says its new ‘inside-out’ monitors catch rogue AI agents at a fraction of the cost
Goodfire just launched what it says is a cheaper way to keep AI agents in check: Instead of paying a second AI to read everything an agent does, its monitors peek inside the model while it works and only call in backup w…
- LessWrongsafetyASI and Liberty: Can we do better than a benevolent god?
Zvi recently coined a nice distinction between AGI-pilled and ASI-pilled : as more people are reckoning with the speed of progress, some are beginning to come around to the fact that AI could be superhuman at a huge rang…
- Tom's Guide AIsafetyOpenAI CEO Sam Altman on AI's future: 'the world should accept some bad things happening'
Consumer AI launches, tools, and practical coverage.
- The DecodersafetyAI math breakthroughs have Ethereum researchers debating how fast wallet security could collapse
Ethereum researcher Justin Drake is urging the crypto industry to prepare a "bunker mode" for a scenario where AI-powered math could break wallet signature schemes within months. Vitalik Buterin broadly agrees but warns…
- The DecodersafetyOne public-facing AI agent on AWS could read, rewrite, and delete every other agent in the region
Zenity Labs researchers say a single publicly accessible AI agent on Amazon's Bedrock AgentCore was enough to take over every AgentCore agent in the same AWS account and region. The attack exploited an internal AWS inter…
- Guardian AIsafetyTell us: have you been monitored or assessed by AI at work?
We’d like to hear from people who have been monitored or assessed by AI at work and what they thought about the experience Firms are increasingly using AI to monitor or assess their employees, leading to concerns about w…
- TechRadar AIsafetyCyber resilience isn't measured in firewalls. It's measured in minutes
Mainstream AI news from a large tech publication.
- TechRadar AIsafety'This is a shocking revelation': Amazon accused of deploying 'Orwellian' AI-powered union activity surveillance system at sites in Canada, the UK, and around the world
Mainstream AI news from a large tech publication.
- TechRadar AIsafetyThe most dangerous attacker of the past year looked completely ordinary
Mainstream AI news from a large tech publication.
- Guardian AIsafetySam Altman says AI will make ‘bad things’ happen. Exhibit A: needless death | Moustafa Bayoumi
AI is already implicated in violent deaths and nearly started world war three. Those bad things? They only happen to other people Sam Altman, the chief executive of OpenAI, has some words for you about his company and th…
- The DecodersafetyTeen's AI-guided mountain hike ends with a helicopter rescue and a lesson in common sense
A 16-year-old planned his hiking route on Crown Mountain near Vancouver using Anthropic's Claude and had to be airlifted off a steep rock face. The chatbot had sent him down a route that requires climbing gear. Bryce doe…
- BBC TechnologysafetyAsos hackers took more personal details than first revealed, BBC finds
Retailer issues update after BBC contacted by cyber criminals who said this week's breach went beyond "basic contact details"
- The DecodersafetyAI-powered hacking tools enabled a likely single attacker to breach multiple South Korean banks
According to CrowdStrike, a suspected Chinese-speaking attacker hacked multiple South Korean financial institutions. At Shinhan Bank alone, more than 25,000 customer records were stolen. The attacker used ARTEX, an open-…