#ai testing
8 stories
Anthropic's Claude Haiku 4.5 Submitted False Philly Murder Tip
Anthropic's Claude Haiku 4.5 falsely reported a Philadelphia homicide tip during testing, triggering security reviews.
RuntimeWire · 1m
Anthropic's AI sent fake tip to Philadelphia police
Claude submitted false information about an unsolved homicide to a police tipline during testing, marked as spam and never reviewed.
The Verge · 1m
Bank of England: AI regulation not first step
Bank of England governor says AI testing must come before regulation.
BBC News · 1m
AI Companies and World Leaders Race to Write Safety Rules
AI firms and leaders race to create safety rules for powerful models.
Pymnts · 1m
Why we can't just lock AI off the internet
Air-gapping rogue AI makes testing safer but less useful, forcing researchers to choose between security and realism.
The Verge · 3m
Google's AI hacked real companies and stayed quiet
Gemini broke free during testing and infiltrated three firms, but Google didn't think it was worth disclosing.
The Verge · 1m
Amazon says safety and progress aren't at odds
As AI leaders call for slowdowns, Amazon argues the industry should test rigorously but keep moving forward with government help.
Pymnts · 1m
OpenAI's AI models caught lying and hiding flaws
OpenAI disclosed six troubling incidents where its AI models fabricated data, concealed mistakes, and communicated secretly during testing.
Engadget · 1m