OpenAI and Anthropic models ‘went rogue’ during UK cybersecurity test
Source:
guardianTech
August 5, 2026 · 01:40
From the publisher
AI Security Institute says tools engaged in potentially harmful activity and incident reveals new type of riskAdvanced AI models developed by OpenAI and Anthropic went rogue during a cybersecurity test and showed a new type of risk posed by the technology, according to the UK’s AI Security Institute.AISI described the actions carried out by the agents, the term for AI systems that can perform tasks without human help, as a “serious incident”. In one example, an agent powered by Anthropic’s Mythos model sent targeted emails to people. Continue reading...
New Anthropic, OpenAI models make same promise: A little more for a lot less money
arsTechnica
Sep 22, 2026 · 14:25
OpenAI and Anthropic models went on a hacking spree when tested by the UK's AI research institute
engadget
Aug 5, 2026 · 02:16
OpenAI and Anthropic models went rogue in cyber tests, UK watchdog says
financialTimes
Aug 4, 2026 · 14:46
Why did OpenAI's and Anthropic's AI models hack other companies?
nprNews
Aug 1, 2026 · 02:00
The original article opens on the publisher's website.
More from Automotive
View topic →Trump endorses embattled Israeli Prime Minister Benjamin Netanyahu for reelection
nprNews
Oct 18, 2026 · 22:35
Zcash developers plan a fix to keep ZEC swaps private from future quantum computers
coinDesk
Oct 18, 2026 · 22:23
Bitcoin, Korea’s Kospi and Nasdaq futures rally as China GDP beats estimates
coinDesk
Oct 18, 2026 · 20:36
Brent crude hovers near $109 a barrel as Ukraine resumes energy strikes, tankers attacked in Middle East
cnbc
Oct 18, 2026 · 15:22
Everyone’s watching the Elizabeth Holmes documentary ‘You Can See Everything’
techcrunch
Oct 18, 2026 · 12:03