OpenAI înăspreşte măsurile de securitate pentru noul model de inteligență artificială Astra - StartupCafe

OpenAI has suspended some internal activities related to its AI model Astra due to concerns about its potential autonomous cyber-attack capabilities. The company indicated that Astra might have reached a 'Critical' risk level, allowing it to attack protected systems without detailed instructions. As a precaution, OpenAI is implementing stricter security controls and will test Astra in isolated environments with enhanced monitoring. Recent incidents involving AI models from OpenAI, Anthropic, and Meta have raised security concerns, prompting regulatory pressures in the U.S. and Europe for more oversight and control mechanisms.