AM-IA brings together news, research papers and applications relating to artificial intelligence published around the world into a single feed — and makes them available in your European language at the click of a button.
Scroll to browse · Click to filter
In a few weeks, Anthropic's Frontier Red Team has sent Claude about 180 million lines of evidence in Rocq, a formal audit work so far reserved for small programs.
Rather than chasing the flaws one by one, Anthropic has Claude map the unnecessarily exposed code. Out of more than 5 million lines, 493 problems have been identified.
Researchers at Anthropic ran a model with 26 billion parameters on 15 $55 set-top boxes and measured the ability of small models to propagate independently from machine to machine.
Preserved reasoning, a two-stage classifier, model partitioning: Anthropic details the arsenal deployed against laboratories that train their models using Claude’s responses.
After identifying more than 29,000 potential vulnerabilities but sorting only 6,000 by hand, Anthropic offers volunteer maintainers the opportunity to receive raw reports from its models directly.
Electrical networks, water, transports: Anthropic joins CrowdStrike, Palo Alto Networks, Dragos or Deloitte to correct faults with Claude, and launches a free open source scanner.
Defence, red teaming or specialised access: verified security teams are granted less restricted access. Glasswing claims to have identified over 129,000 confirmed vulnerabilities.
Anthropic has evaluated Z.ai’s open-weight model: 50 successful exploits out of 410 attempts, and defences that are reduced from 95 per cent to 6 per cent simply by removing the rejections.
The September report identifies espionage, fraud and other abuses that Anthropic interrupted between December 2025 and August 2026.