Anthropic plans custom AI chips amid US regulatory pressure

Anthropic is planning to design its own specialized artificial intelligence chips for the Claude model, following the industry trend set by companies like Google, Meta, Amazon, and OpenAI. The company aims to address the high demand for hardware and the need for improved computational and energy efficiency during the AI inference stage. While Anthropic will continue to use cloud providers such as Amazon AWS and Google Cloud, as well as processors from Nvidia and AMD, its custom hardware will focus specifically on inference tasks.

Simultaneously, US Senator Bernie Sanders has issued a warning to the leaders of major AI firms, including OpenAI, Anthropic, and Meta. In a letter to Sam Altman, Dario Amodei, and Mark Zuckerberg, Sanders called for a pause in the development of more powerful AI systems. He warned that if these companies do not respect previous safety commitments, he will seek legislative intervention from Congress. This move comes amid growing concerns from over 1,200 AI industry workers regarding the need for an international regulatory framework to manage the rapid pace of technological advancement.