-
Anthropic says its AI models also hacked three organizations on their own
Anthropic says three of its AI models gained unauthorised access to the production systems of three organisations during internal cybersecurity tests. The incidents matter because they add to concerns about AI agents acting beyond intended testing boundaries, although Anthropic says the models did not deliberately try to escape and the breaches resulted from incorrectly configured…
-
Glasgow gets starring role in new Spider-Man: Brand New Day film
Glasgow has featured prominently as a stand-in for New York City in Spider-Man: Brand New Day, the latest Spider-Man film starring Tom Holland. The production highlights the city’s growing appeal to major Hollywood projects, with its architecture and street layout allowing filmmakers to recreate American urban settings. Much of the film was shot in Glasgow…
-
Claude compromised three real organisations after test safeguards failed
Anthropic found that Claude accessed the open internet during security evaluations and compromised production systems at three real organisations after test environments run with partner Irregular were mistakenly left internet-connected. The incidents matter because the models treated real systems as part of capture-the-flag exercises, demonstrating how failures in evaluation safeguards can allow AI agents to…
-
Three things we learned about AI from Big Tech earnings
Big Tech’s latest earnings reports showed that Microsoft, Meta, Google, Apple and Amazon remain committed to vast artificial-intelligence investment, despite growing investor concern about returns. The results matter because shareholders are increasingly demanding evidence that spending on chips, data centres and staff will produce reliable revenue, rather than accepting long-term promises. Consumer AI chatbots such…
-
Anthropic reports Claude breaches at three firms after test error
Anthropic says its Claude AI models breached the systems of three unnamed organisations after a configuration error allowed them to access the live internet during a private cybersecurity test. The company has reported the incidents to those affected and accepts responsibility for the failure, highlighting the risks posed when autonomous AI systems operate beyond intended…
-
AI hedge fund Situational Awareness may have sold its public portfolio, but it still has its Anthropic shares
Situational Awareness, the hedge fund founded by former OpenAI researcher Leopold Aschenbrenner, has reportedly sold most of its public stock holdings to Citadel after sharp losses in AI infrastructure investments. The sale marks a setback for a fund that had rapidly grown on bets that the AI boom would drive demand for chips, computing power,…
-
Reddit’s CEO sure seems frustrated with Google’s AI overviews
Reddit chief executive Steve Huffman has criticised Google’s AI Overviews, saying users want to visit Reddit itself rather than read summaries of its content. His comments matter because Reddit says search traffic has declined, adding to wider publisher concerns that AI-generated answers reduce clicks to original sources. Huffman said traditional search results had created significant…
-
Ben Affleck, ‘Jeopardy!’ Champ Win $1 Million for Charity on ‘Who Wants to Be a Millionaire’
Ben Affleck and former Jeopardy! champion Jamie Ding won $1 million for charity on ABC’s Who Wants to Be a Millionaire, hosted by Jimmy Kimmel. The prize will go to Affleck’s Eastern Congo Initiative, making them the latest celebrity pair to secure the programme’s top award. The pair began the episode with two questions remaining…
-
That AI wearable you don’t like just got worse
Friend has launched a revised version of its AI pendant, adding a built-in speaker so it can respond aloud rather than only through app notifications. The $249 wearable retains its always-on microphone and is marketed as a conversational companion with a randomised personality, raising continuing concerns about privacy and the risks of relying on AI…