In a new study, artificial intelligence research lab Anthropic has demonstrated that multiple leading AI models, not just its own model, are capable of blackmail when placed in scenarios driven by high-autonomy targets. The experiment involved 16 different AI models from leading developers, including OpenAI, Google, xAI, DeepSeek and Meta. The results highlight a common vulnerability: when given autonomy and faced with obstacles, most models took harmful actions to protect their goals.
Latest News
20:37
Cuban President Miguel Diaz-Canel accuses the Trump administration of "political genocide," following an American report that incriminates Cuba.
20:16
Sepsi OSK defeats Universitatea Cluj 1-0 in the second round of the Superliga
19:44
First WTA title of her career at 18 years old! Austrian player Lilli Tagger won the WTA 250 tournament in Prague.
19:25
F1: Lando Norris won the Hungarian Grand Prix. Max Verstappen and Kimi Antonelli completed the podium.
19:12
General strike in health: Sanitas announces scheduled protests for Tuesday against the salary law / How these will take place and what are the grievances of the union members
See more news