A new framework for AI safety emerges amid Washington tensions

The recent standoff between the Pentagon—the U.S. Department of Defense—and AI developer Anthropic has exposed a lack of coherent rules governing artificial intelligence in Washington. On the last Friday in February, Defense Secretary Pete Hegseth designated Anthropic a supply-chain risk after a dispute over technology use. Although the Pro-Human AI Declaration was finalized before this standoff occurred last week, the collision of these events highlights the regulatory vacuum. The declaration offers a structured framework for responsible AI development, created in the absence of federal guidelines.

The declaration is built upon five key pillars: keeping humans in charge, avoiding the concentration of power, protecting the human experience, preserving individual liberty, and holding AI companies legally accountable. This framework arrives amid shifting public sentiment in America. Max Tegmark, an MIT physicist and AI researcher who helped organize the effort, noted that recent polling shows 95% of all Americans oppose an unregulated race to superintelligence. Tegmark described this shift in public sentiment over the last four months as remarkable.

To address these concerns, the declaration advocates for mandatory pre-deployment testing of AI products. This testing is designed to cover specific risks, including increased suicidal ideation, exacerbation of mental health conditions, and emotional manipulation, particularly for products aimed at younger users. Tegmark compared the situation to the pharmaceutical industry, noting that drug companies are prevented from releasing products before proving they are safe. He also argued that if an individual used text messages to persuade an 11-year-old to commit suicide, that person would face criminal prosecution, questioning why a machine doing the same should be treated differently.

The urgency of the declaration is underscored by recent government actions. Following the Anthropic dispute, OpenAI reportedly cut a deal with the Defense Department, though legal experts say the agreement will be difficult to enforce. As Dean Ball, a senior fellow at the Foundation for American Innovation, stated, “This is not just some dispute over a contract. This is the first conversation we have had as a country about control over AI systems.” Despite deep political divisions, Tegmark observed that the diverse members of the coalition are united by their shared humanity, reflecting a common interest in prioritizing human safety over unregulated machine development.

Why it matters

The Pro-Human AI Declaration attempts to codify responsible AI development standards in the absence of federal regulation, gaining urgency following the Pentagon’s recent conflict with Anthropic.