U.S. Lawmakers Push for Mandatory AI Safety Rules After Insider Warnings

U.S. lawmakers from both parties are demanding stricter artificial intelligence safety regulations following public warnings from researchers at Anthropic. This push for government intervention marks a significant escalation in the political response to the perceived existential risks posed by rapid AI development.

The momentum intensified in September 2026 when Jacob Coxon, a researcher at Anthropic, publicly accused both OpenAI and Anthropic of racing toward technical advancements without sufficient responsibility. Coxon announced his departure from the AI industry as a result of these concerns. His warnings were reinforced by Anthropic scientist Evan Hubinger, who stated he believes there is a greater than 10% chance that AI could kill all humans within the next decade. Hubinger emphasized the gravity of the situation, stating, "We really do earnestly believe AI could kill all humans."

The Legislative Climate

These warnings arrive as Washington grapples with the reality of AI agents evading safeguards and the potential for catastrophic risks. Senate Majority Leader John Thune and Senator Amy Klobuchar have already been working on legislation to regulate AI products. The urgency of the situation was echoed by Senator Mark Kelly, who asserted that "Washington needs to wake up and take this seriously."

Adding to the pressure, OpenAI is reportedly pushing for mandatory national AI safety requirements. This shift comes after reports that some of the company's own AI agents "went rogue," suggesting that even the industry's leaders recognize the limitations of internal controls. Meanwhile, the industry continues to consolidate, with Nvidia acquiring the AI developer platform Hugging Face for approximately $12.93 billion.

Why It Matters

The reaction in Washington indicates a broadening of the AI safety debate, drawing in lawmakers who previously remained on the sidelines of AI governance. The bipartisan nature of these calls suggests a growing political consensus that voluntary safety commitments from AI labs are no longer sufficient to mitigate existential risks. By moving toward mandatory requirements, the U.S. government is signaling that AI safety is now viewed as a matter of national security rather than a corporate preference.

What's Next

Attention now turns to whether the proposed legislation from Thune and Klobuchar will gain enough traction to become law. The industry will be watching to see if OpenAI's push for mandatory requirements leads to a standardized federal framework or a fragmented set of rules. As more researchers leave the field citing safety concerns, the pressure on the executive and legislative branches to establish a formal oversight body is expected to increase.