AI liability fight reaches Senate as Google releases powerful new model

Key Points:

• Chinese hacking group TA419 has targeted prominent U.S. AI experts and institutions with phishing campaigns since at least 2025, impersonating figures like former White House official Lynne Edwards Parker to steal passwords.
• The hackers used fake invitations to join AI policy committees or contribute to Senate reports, sending malicious links to gain access to victims' cloud accounts upon response.
• Broadcom plans to provide up to $42 billion in financing to AI company Anthropic following its upcoming IPO, which could value the company at $2 trillion, with options to convert debt into shares.
• The UK’s MI5 issued an espionage alert accusing China's CGTRI of funding UK academic research to enhance Chinese Ministry of State Security capabilities, urging universities to scrutinize collaborations and funding sources.
• UK security officials have called for ending ties with Chinese organizations due to espionage concerns, while Chinese authorities deny all allegations of spying.