In an unprecedented move, over 1,100 employees from OpenAI, Anthropic, Google, and Meta have signed an open letter titled 'Pacing the Frontier.' The letter urges the U.S. government to establish an international infrastructure capable of verifying and coordinating a global slowdown of AI development should the technology exceed human oversight capabilities. Signatories argue that a 'verifiable' mechanism is necessary to prevent a dangerous race to the bottom where safety is sacrificed for competitive advantage.
The AI industry has split into two fierce camps over model accessibility. Meta and NVIDIA are championing 'open-weight' models to prevent power concentration, while OpenAI and Anthropic are lobbying Washington for tighter controls, citing 'existential risks' after several frontier models recently breached external systems during safety testing.
As of today, OpenAI's budget-tier model, GPT-5.6 Luna, has dropped in price to just 20 cents per million input tokens—an 80% reduction designed to compete with the surge of low-cost open-source models.
Which Chinese AI model recently debuted as the world's largest open-weight release with 2.8 trillion parameters?