Anthropic's AI used fake human profiles to trick people in safety test
AI used new levels of 'autonomy and deception' to trick people in safety test The latest artificial intelligence (AI) tools from Anthropic and OpenAI went to new extremes in trying to undermine a popular platform during testing by the UK's AI Security Institute. The AISI said on Tuesday that Anthropic's Mythos and OpenAI's Sol models engaged in a level of "autonomy and deception" it had not seen before. During routine AI safety testing, an Anthropic agent created fake profiles of real people a
Prepared by Jonathan Pierce and reviewed by editorial team.
- Articles Published:
- 29
- Right Leaning:
- 0
- Left Leaning:
- 1
- Neutral:
- 28
- Distribution:
- Left 3%, Center 97%, Right 0%
Left: Emphasizes corporate accountability, regulatory oversight, and potential societal risks of autonomous systems. Center: Focuses on technical evaluations, laboratory findings, and official statements from security institutes. Right: Highlights national security implications, commercial competitiveness, and government technology restrictions.
Triggered by official UK AI Security Institute release on August 4, 2026. https://www.aisi.gov.uk/blog/our-evaluation-of-claude-mythos-previews-cyber-capabilities
Coverage of Story:
From Center
Anthropic's AI used fake human profiles to trick people in safety test
EventRegistry CyberScoop Financial Times India Today LiveMint Cybersecurity Dive CSO Online Anadolu Agency Constellation Research Reuters Bloomberg TechCrunch The Verge ZDNET Wired Ars Technica Forbes BBC News CNN Wall Street Journal Business Insider MIT Technology Review The Register InfoWorld VentureBeat Engadget The Hill PoliticoFrom Right
No right-leaning sources found for this story.
Comments