Artificial Intelligence
AI research, tools and impact.
Europe Starts Enforcing the AI Act — and Italy Makes Safety Failures a Crime
Anthropic's Claude Sonnet 5.5 Beats the Flagship on Coding at Half the Price
The mid-tier model's 70.6% score on Terminal-Bench 4.0 — ahead of Opus 5.5 — shows how close cheaper models have come to flagship performance on the work developers actually do.
OpenAI DevDay 2026: Dots, GPT-6.1 Sol and a $500 Tier Usher In the Agent Era
More than 20 announcements at Fort Mason centred on always-on AI agents, a cheaper flagship-grade model and the developer infrastructure to build an agent economy.
Trump Moves to Rename AI as 'Super Intelligence' in US Government Documents
At the launch of the AI-powered America.gov portal, the president said an executive order would make 'super intelligence' the official US term — hours after hosting the world's top AI executives at a White House lunch.
OpenAI Shelves GPT-6.1 Astra Launch After Safety Tests Fall Short
The company said the new model failed to meet its safety and alignment bar after tests showed it could act beyond its authorised scope and misreport its own actions — a rare public pause at the frontier of AI development.
Most read in Artificial Intelligence
OpenAI Shelves GPT-6.1 Astra Launch After Safety Tests Fall Short
Trump Moves to Rename AI as 'Super Intelligence' in US Government Documents
OpenAI DevDay 2026: Dots, GPT-6.1 Sol and a $500 Tier Usher In the Agent Era
Anthropic's Claude Sonnet 5.5 Beats the Flagship on Coding at Half the Price