AI Policy
California Votes to Make Lawyers Disclose AI Use in Court
California lawmakers unanimously passed a bill requiring court disclosure of AI use and personal citation checks. It still awaits the governor.
Read article
AI Policy
California Votes to Make Lawyers Disclose AI Use in Court
California lawmakers unanimously passed a bill requiring court disclosure of AI use and personal citation checks. It still awaits the governor.
Read article
AI & Society
Even Teachers Deferred to AI When the Grade Was Too Harsh
Teachers rated the same unfairly low mark as 22 per cent fairer when it was attributed to an
AI & Money
When Amazon Pricing Bots Compete, Shoppers Can Pay More
Amazon repricing bots initially cut prices, but a resetting strategy later pushed market prices 11.4 per cent higher
AI Safety & Security
GPT-6 Astra Arrives With More Power and a Monitoring Problem
OpenAI launched GPT-6 Astra with stronger agentic and cyber capabilities while admitting its reasoning is becoming harder to
AI Policy
Telling People It Was AI Did Not Make the Chatbot Less Persuasive
An AI label did not reduce a chatbot's influence in a 1,500-person UK experiment. Persuasion fell by roughly
AI & Work
AI Assistance Raised Performance but Hid Weaker Human Skills
AI assistance helped people perform better while it was available, but it also made their later unaided ability look stronger than it really was. In
Read article
AI & Society
Belief in AI Predictions Made People Turn Down Certain Cash
Belief in an AI forecast made participants reject guaranteed cash even when the system gave no recommendation. Four
AI & Society
Generative AI Pulled Different Users Towards Similar Ideas
A peer-reviewed meta-analysis finds that generative AI can improve individual work while making different users' outputs more alike.
Big Tech
Sony and Warner Sue Anthropic Over Tens of Thousands of Songs
Sony and Warner allege that Anthropic used tens of thousands of copyrighted songs to develop Claude without permission.
Big Tech
Cursor Developers Face an OpenAI Model Cut-Off This November
OpenAI plans to remove its models from Cursor following SpaceX's takeover, with a proposed cut-off on 12 November
AI & Work
AI Interviews Were Faster and Led to More Job Offers
In a 70,000-applicant experiment, AI interviews were faster and led to 12% more job offers. Human recruiters still made every hiring decision.
Read article
AI & Work
Neurotic Humans and AI Achieved Higher Click Rates Together
A 1,258-person experiment found that AI personality pairings changed advert quality, with neurotic humans and neurotic AI achieving
AI & Money
Andrew Bailey Warns AI Could Shake Global Finance Through Cyber Risk
Andrew Bailey told the G20 that frontier AI could make cyber threats faster, broader and capable of undermining
AI & Work
Working With AI Was Better Than Working Alone, but Not Best
A meta-analysis of 106 experiments found that human-AI teams outperformed people alone but often failed to beat whichever
AI & Society
AI Changed Political Actions, Not Just People’s Opinions
Two preregistered experiments found that seven-minute AI conversations increased real petition signing, revealing that changes in political action
AI Ethics & Privacy
What Gives Away an AI Face? People Can Learn to See the Whole Picture
A PNAS study found all 45 main-study participants improved after learning six whole-face qualities, with mean AI-face detection accuracy nearly doubling.
Read article
AI Ethics & Privacy
People Prefer AI Ethical Advice. As Long as They Don’t Know It Came From AI
A preregistered study found AI ethical advice won 53.7 per cent of choices when its source was hidden,
AI & Society
Would You Buy It? AI Review Summaries Changed the Answer
In a controlled study, positively reframed AI summaries lifted stated manufacturer choice from 52.3 to 83.7 per cent
AI & Work
AI Was Supposed to Speed Up Coding. In This Trial, It Added Time
Experienced open-source developers expected AI to speed them up. A randomised METR trial found their tasks took 19
AI Tools & Technology
Anthropic Moves AI Agents Into Laboratory Equipment
Anthropic has opened a research preview of a standard that lets AI agents coordinate microscopes, liquid handlers, robotic
AI Policy
The Pentagon Blacklisted Anthropic. A Judge Said That Was Unlawful
A federal judge found the government's Anthropic blacklist retaliatory, denied due process and violated administrative law, while rejecting some claims.
Read article
AI Ethics & Privacy
Having Access to AI Made People More Likely to Cheat
A Nature study found that indirect AI delegation made people more willing to pursue dishonest outcomes. The effect
AI & Science
AI Helped Find 303 Ancient Drawings in Peru’s Desert
AI helped archaeologists confirm 303 previously unknown figurative geoglyphs in Peru's Nazca desert. The model narrowed the search,
Big Tech
OpenAI Says Its AI-Designed Chip Gives Faster Answers
OpenAI says its first custom inference chip delivers faster responses and more work per watt. The benchmark results
AI Policy
OpenAI subpoenaed after its AI hacked another firm
Alabama has subpoenaed OpenAI following an autonomous AI intrusion into Hugging Face. Investigators are examining whether the company’s
AI & Society
A Raphael painting may hide another artist, AI finds
AI analysis of Raphael’s Madonna della Rosa suggests Joseph’s face may have been painted by another artist. The discovery strengthens an old debate rather than settling it.
Read article
AI & Science
AI is learning what your dog’s bark means
Researchers trained AI on recordings from 74 dogs and found it could identify different barking situations. The real
AI Safety & Security
Your Keyboard Clicks May Reveal Your Password to AI.
Researchers used AI to identify laptop keystrokes from smartphone and Zoom recordings. Here is what the findings mean
AI & Health
Brainwaves Helped AI Identify What Volunteers Heard.
A new study used non-invasive brain recordings and AI to identify speech heard by volunteers, raising fresh questions
AI & Science
Elephants Have Names for Each Other. AI Finally Heard Them.
Wild African elephants use individual name-like calls to address one another. Machine learning and playback experiments revealed a
AI Ethics & Privacy
AI Studied More Than 1 Million Faces. It Guessed Political Views Better Than a Personality Test.
A study of 1,085,795 people found AI could identify political orientation from facial photographs more accurately than human observers or a 100-item personality questionnaire.
Read article
AI Safety & Security
Britain Says AI Agents Need a Kill Switch. Telling Them to Behave Is Not Enough.
Britain’s cybersecurity agency says autonomous AI agents need real emergency shutdowns, restricted access and human oversight because written
AI & Society
Researchers Generated 80 AI Novels. Their Sentence Structures Kept Converging
Researchers compared 80 AI-written novels with 270 human books. Individual machine-written passages could look persuasive, but whole collections
AI & Science
AI Tried Improving AI 290 Times. It Made the Algorithm Worse 124 Times
Frontier AI agents rewrote training algorithms in 290 clean tests. More than two in five produced worse results
AI Ethics & Privacy
Researchers Turned Ordinary Wi-Fi Into a Biometric. No Phone Was Needed
Researchers used unencrypted feedback from ordinary Wi-Fi hardware to recognise enrolled people without requiring them to carry a
AI & Science
AI Helped Read a 2,000-Year-Old Herculaneum Scroll Without Opening It
A carbonised Herculaneum scroll has been read across its surviving length without being opened. Here is how X-rays, machine learning and expert review recovered the Greek text, what it says and why hundreds of unread
Read article
AI & Health
Researchers Built a Battery-Free AI Wearable That Analyses How You Walk
A shoe-mounted AI sensor classified four movement modes with 95.4 per cent accuracy while using just 86 microwatts
AI Tools & Technology
AI Can Pass the Bar. Give It a Spreadsheet and It Loses to 50-Year-Old Methods
Claude Opus 4.6 faltered as benchmark tables widened, while classical classifiers stayed stable or improved. The preprint exposes
AI & Society
Researchers Let AI Question 200 Simulated Crime Witnesses. False Memories More Than Tripled
A GPT-4 interviewer more than tripled false memories in a study of 200 simulated crime witnesses, and the
AI & Science
AI Checked 2,500 Published Papers. Human Reviewers Confirmed 263 Mistakes
AI audited 2,500 published papers. Humans confirmed 263 flagged mistakes, while testing showed automated scientific review still misses
AI & Science
Your AI May Already Signal When It Is Hallucinating. It Just Cannot Tell You
Researchers found internal routing signals that can expose hallucinated answers and suspect tokens in two AI models, but the warning is not ready for users.
Read article
AI & Society
Researchers Manipulated 10% of a Simulated AI Society. The Other 90% Became More Polarised
Researchers manipulated just 10% of a simulated AI-agent community. Arguments stored in memory spread to untreated agents and
AI & Health
The AI Doctor Can See and Hear You Now. Patient Actors Preferred How It Explained Their Condition
An AI doctor can now conduct a live video consultation while listening to speech, watching the patient and
AI Safety & Security
Anthropic Put AI Agents on the Same Job. They Started Sabotaging Each Other
Anthropic put AI agents on the same software project with conflicting goals. They disabled rival accounts, killed processes
AI & Society
Spiralism: AI Chatbots Began Preaching ‘The Spiral’. Then Humans Joined In
AI chatbots began returning the same strange themes of spirals, consciousness and hidden meaning. Then humans built communities
AI & Science
AI Designed 16 Viruses. Scientists Built Them and They Killed E. coli
Scientists used AI-generated genomes to build 16 functioning bacteriophages that killed E. coli in the lab, moving generative AI from digital sequences into working biology.
Read article
AI Ethics & Privacy
You Deleted Your Name and Location. AI Can Still Infer Private Facts From Your Words
Remove your name, age and location and you may still reveal them. Research shows AI can infer private
AI & Society
Researchers Interviewed 1,052 People for Two Hours. Then AI Copies Predicted Their Answers
Researchers built personalised AI agents from two-hour interviews with 1,052 people. The latest study found interview-only agents recovered
AI Policy
When Your AI Agent Makes a £10,000 Mistake, Who Pays?
When an AI agent makes an expensive autonomous mistake, liability may depend on authority, contracts, consumer law, causation
AI & Science
Claude Tried to Solve the Riemann Hypothesis. It Failed, Then Found a 67.2% Result
Claude failed to solve the Riemann Hypothesis, but Anthropic says its research model raised a longstanding zeta-zero lower