Big Tech’s AI safety rift signals disruption and disparity for enterprises

Major AI companies disagree on how to secure powerful AI models, creating unpredictable access and deployment conditions for businesses rather than industry-wide slowdowns. Companies are applying different safety approaches, release schedules, and usage restrictions, meaning enterprises may access the same AI capabilities at different times and under different rules. An emerging "AI assurance" layer (third-party evaluations of models for safety and compliance) is developing, but enterprises should not assume a single evaluation means an AI system is fully safe.