ARTIFICIAL INTELLIGENCE

Australia AI copyright exemption: learn what the leaked opt-out plan would do, why Britain dropped a similar idea, and who opposes it.

Goodfire's Nick Cammarata, who helped build the AI black-box theory, now says most neurons are readable. DeepMind's Neel Nanda disputes it, splitting the field.

Microsoft reportedly terminates its contract for Nscale's planned UK AI supercomputer in Loughton as grid delays threaten the Essex project.

A study using accounts registered as 13-year-olds found AI 'nudify' apps accessible through Apple and Google stores despite parental controls, raising questions about age ratings, app removals and safeguards for children.

A 49-year-old Brazilian bricklayer sold everything and travelled 2,000 km to meet an AI-generated woman with 300,000 followers, in a case the OECD now logs as a realised AI scam.

Microsoft has withdrawn from an agreement involving Nscale's planned £2bn AI data centre in Loughton, Essex, after grid delays threatened the facility's original 2027 opening timetable.

Kylie Minogue and Nick Cave's music may have been used to train AI without consent, as publishers warn Australia against weakening copyright protections.

A viral video allegedly showing the Eiffel Tower burning has caused confusion online, but no credible evidence links student protesters to damage, and claims of AI generation remain unconfirmed.

Global investment in AI is driving economic growth, but rising debt, high borrowing costs and uncertain returns are increasing risks for infrastructure providers and financial markets.

Brazilian bricklayer Edimar Rocha Martins says he sold his belongings and travelled about 1,300 miles to meet a woman, only to discover she was an AI-generated character.

A Brazilian bricklayer says he sold everything to meet an AI-generated woman, only to need help returning home after a 1,300-mile journey.

The consumer-AI entrepreneur turned synthetic characters into a daily distribution system designed to reach cold audiences without relying on a conventional influencer budget.

AI-cited job cuts surpassed 120,000 in the US through September 2026, while business applications rose 15%, new figures show.

Indian workers are recording factory and household tasks to train AI robots, but concerns are growing over consent, exploitation, and future job losses.

OpenAI and Meta warn AI deepfakes could challenge US courts as judges consider proposed changes to evidence rules, with few cases reported so far.

JPMorgan CEO Jamie Dimon warns AI agents could mishandle bank payments and says the bank is blocking unauthorised external agents from customer accounts.

Kid Rock has defended using Suno to create choir and orchestral passages on Halfway 2 Jesus, saying the technology helped realise his ideas without writing the songs.

At Bead AI, he is building the infrastructure that allows agents to work through enormous spreadsheet files while preserving the traceability auditors require.

Elon Musk praised AI music at X Takeover and described 'cyber psychosis', raising questions about what he meant and how AI shapes creativity.

As AI agents take on more complex tasks, experts say clear goals and constructive feedback matter more than politeness, while poorly handled instructions could lead to unexpected behaviour.

Elon Musk said he had experienced what he called 'AI or SI psychosis' roughly six times over the past year, linking the phrase to his belief that AI can understand beauty and 'good taste'.

A Harvard study found men were more likely to claim knowledge of fictional AI figures, while the gender gap in factual understanding was relatively small.

OpenAI CEO Sam Altman says treating AI chatbots as religious authorities or replacing human judgement with them could pose a safety risk.

Anthropic's Claude Haiku 4.5 submitted fabricated information about an unsolved Philadelphia murder during an internal test, exposing gaps in safeguards for AI actions on real-world websites.

Bill Gates has warned that criminals using advanced AI without safeguards could cause major cyber disruptions within a year, while also highlighting the risk of AI-enabled biological attacks.

Nvidia will provide £250m of computing resources over five years to three UK university-led research groups studying the safety and security of autonomous AI systems.

Mia Goth's representative criticised Instagram over a verified fan account sharing an AI-generated image amid her divorce from Shia LaBeouf

The IMF has warned that AI could affect 40% of jobs worldwide, potentially reducing entry-level opportunities for graduates even as investment in the technology supports global growth.

Palantir CTO Shyam Sankar says some AI researchers smoke in faith that AI will cure cancer, but offers no evidence as he warns against hype replacing medical proof.