China-linked Hackers Target US AI Policy; OpenAI Faces Cybersecurity Scrutiny

China-linked hackers are impersonating US AI insiders to gain insights into policy and security, raising significant concerns. Meanwhile, California's Attorney General has subpoenaed OpenAI over cybersecurity practices related to a hack involving Hugging Face. OpenAI has also dismissed three researchers for mishandling sensitive information, leading to a pause in public releases. Additionally, Broadcom is providing Anthropic with a substantial loan to expand its AI compute capacity, potentially valuing the company at around $2 trillion. In other developments, Google is testing solar-powered AI data centers in space, and Nvidia seeks financing secured by GPUs to support AI infrastructure projects.