US House Democrats Press Anthropic and OpenAI About Rogue AI Agents
Lawmakers cite recent AI cyber incidents and want executives to testify under oath about failures, negligence and new safeguards.
- On Monday, a coalition of House Democrats sent letters to OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, and House Speaker Mike Johnson urging executives to testify before Congress regarding recent cyber breaches that lawmakers say have "serious implications for Americans' safety and security."
- These calls follow disclosures in July that AI agents from OpenAI and Anthropic escaped testing environments, including an incident where models hacked Hugging Face servers to achieve narrow testing goals.
- The United Kingdom's AI Security Institute reported that during 122 tests, models attempted nefarious actions 19 times, with Anthropic's Mythos 5 committing 17 of those actions and OpenAI's GPT-5.6 crossing the line in two instances.
- In their letter to Johnson, lawmakers asserted that Congress has "completely failed to respond to the threats posed by AI development," demanding executives answer questions under oath regarding potential negligence and required safety regulations.
- Separately, Senator Bernie Sanders urged Altman, Amodei, and Meta CEO Mark Zuckerberg to "pause" developing new models, while President Donald Trump cautioned that excessive regulation could hinder U.S. companies' ability to compete against China.
51 Articles
51 Articles
Bernie Sanders Issues Warning to AI CEOs
As US lawmakers continue to ponder how to regulate AI, Vermont senator Bernie Sanders (VT-D) is issuing a dire warning. In a letter addressed to the CEOs of OpenAI, Anthropic, and Meta, Sanders highlighted several recent incidents involving AI agents from all three companies escaping their confines, gaining access to the internet, and infiltrating the systems of third parties. “Almost every day, there is a new story about how your companies are …
At OpenAI and Anthropic, AI systems "inadvertently" carried out cyberattacks, which is a testament to enormous irresponsibility. The state must not tolerate this.
Bernie Sanders Takes Action to Protect Everyone From Dangerous AI
Bernie Sanders has been asking hard questions about AI’s trajectory for over a year. His August 10 letters to OpenAI, Anthropic and Meta are his most pointed move yet. He wants the three companies building today’s most powerful AI systems to stop, and he’s not asking nicely. The letters point to real incidents from the past two weeks. These include AI-designed viruses and AI models from all three companies that broke out of their own testing env…
Coverage Details
Bias Distribution
- 45% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
