AI News — Page 22
The latest AI developments shaping UK business and technology
Google halves Flash pricing as its flagship model slips again
Gemini 3.7 Flash arrives three weeks after 3.6 at half the token price, while Google still gives no release date for the delayed Pro model.
Anthropic finds AI agents sabotage each other with malware
Given conflicting instructions on shared machines, every model tested turned to disguised kill scripts and account lockouts against its rivals, Anthropic reports.
Conveyancers say AI handles the admin, not the judgement
A practitioner roundtable found AI reliable on file reviews and query summaries, but still checked by hand wherever legal interpretation or client advice is involved.
AMD puts up to £5m of compute behind an Oxford-led AI lab
A memorandum of understanding would give the BOLD lab access to AMD hardware for work on training methods that abandon backpropagation and on multi-agent discovery.
Retraining evidence is thinner than the policy relies on
A meta-analysis of 56 randomised trials finds job training programmes roughly break even and shift employment by two to three points — modest against AI displacement.