OpenAI's decision to halt some internal activities regarding its unreleased model Astra stems from fears that it may possess 'Critical' capabilities, potentially allowing it to autonomously launch cyberattacks. This concern follows recent security incidents involving AI systems from other companies, including Meta and Anthropic, which have raised alarms about the safety of AI technologies.
In response to these incidents, U.S. lawmakers are advocating for the 'AI Kill Switch Act,' which would mandate that AI companies retain the ability to deactivate or limit their models to prevent unauthorized actions. Representative Ted Lieu emphasized the urgency of passing this legislation, citing unauthorized hacks by advanced AI models.
Additionally, the White House and the European Union are actively working on frameworks and regulations to oversee AI development and deployment, indicating a broader push for accountability in the AI sector