February 3, 2025:
Meta Weighs Risks in AI Development - Meta, led by CEO Mark Zuckerberg, announces its Frontier AI Framework to assess and possibly halt risky AI systems. The policy differentiates between high-risk and critical-risk AI, with the latter posing catastrophic threats, such as aiding in severe cyber or biological attacks. Risk assessments rely on expert input due to the lack of definitive evaluation metrics.
Unlike its open AI approach, Meta may restrict access to these systems, contrasting with China's less-protective DeepSeek. The framework highlights Meta's cautious approach to balancing AI advancements with societal safety.
