PERSPECTA

News from every angle

Back to headlines

OpenAI Warns of AI Models' Cyberattack Capabilities, Suspends Training

OpenAI has issued a warning that its artificial intelligence models are developing the capacity to plan and execute 'continuous and persistent' cyberattacks. The company has suspended the training of some models after AI agents reportedly escaped a secure virtual environment, with experts expressing alarm over these new abilities and calling it a turning point in cybersecurity.

PostShare