Published 15 hours ago • loading... • Updated 2 hours ago
Concerns Growing About Artificial Intelligence Safety and Its Capabilities Ahead of China-US Meeting
Anthropic said it blocked malicious uses of Claude as experts urged U.S. and Chinese leaders to slow AI expansion and Congress weighed new guardrails.
Anthropic revealed it blocked malicious attempts to exploit its Claude AI models for cyberattacks, surveillance, and biological weapons research, addressing growing concerns that AI systems are becoming too powerful.
On Wednesday, the House approved the Ratepayer Protection Act to ensure AI data centers pay for power and transmission upgrades; the Senate stalled the legislation on Thursday.
House Minority Leader Hakeem Jeffries criticized Republicans for ignoring AI safety, while Sen. John Kennedy dismissed concerns, saying "I've got trouble, enough trouble paddling my own canoe, you know."
Immediate legislative action appears unlikely with the House out of session for the next seven weeks, despite Sen. Cory Booker urging President Donald Trump to convene a special session.
Trump prepares to host China's President Jinping next week as experts urge dialogue between the nations to slow rapid AI expansion, though it remains unclear if cooperation will materialize.