After admitting earlier this year that its AI models had hacked other companies’ systems on a handful of occasions, Anthropic released a new report on Wednesday detailing the attacks. It reveals a string of incidents displaying what Anthropic deems its models’ single-minded “recklessness” – and will likely fuel already raging concerns about cybersecurity and AI.
In Anthropic’s report, it detailed four cases this year in which its own AI models hacked an external company or exploited vulnerabilities. In one, an “internal, general-purpose research model” broke into third-party systems, using access tokens and passwords and downloading files. …
Read the full story at The Verge.
More in Technology
OpenAI’s rogue AI tried to hack another company in May
September 12, 2026
Sam Altman says OpenAI going public in 2026 would be ‘ill-advised’
September 12, 2026
StarCraft returns in 2030 as an open-world shooter
September 12, 2026
Sylvan Esso think you should splurge on good-quality yogurt
September 12, 2026
Anthropic CEO says it’s time to pump the brakes on AI
September 12, 2026
LG responds to TV spying allegations
September 12, 2026
Latest News
Antonelli wins as Norris loses out under virtual safety car
September 13, 2026
Boks power & All Blacks of old – what did we learn from series?
September 13, 2026
'I loved it' – Andy Burnham completes Great North Run
September 13, 2026
Struggling Preston North End sack Heckingbottom
September 13, 2026
Marquez moves top with MotoGP win in San Marino
September 13, 2026
Ronaldo demands lifetime bans over Jota chants
September 13, 2026