Artificial intelligence is dramatically lowering the cost and technical skill required to develop biological weapons, creating what national security experts describe as an emerging crisis that demands immediate U.S. action.
Developing bioweapons has become far cheaper and easier over the past century. AI now compounds that risk by automating critical steps in the weapons development process, according to a new analysis circulating among policymakers.
Defending against AI-enabled bioweapons requires multiple layers of defense. No single policy fix will work. U.S. officials must focus on critical chokepoints: safeguards for AI models used in biological research and stronger oversight of nucleic acid synthesis—the process of manufacturing genetic material that could be weaponized.
Those safeguards face real obstacles. The speed of AI advancement, the diversity of biological threats, and the difficulty of enforcement mean no easy solution exists.
The Trump administration's July 2025 AI Action Plan names biosecurity investment as a core national security priority—a sign the White House recognizes the threat is real.
Historically, the high cost of building weapons of mass destruction protected the country. Building a nuclear bomb required billions of dollars and world-class scientists. Saddam Hussein's Iraq spent roughly $10 billion on its nuclear program and never succeeded. That economic barrier mattered.
But biological weapons are different. MIT professor Kevin Esvelt, testifying to Congress in 2022, warned that the world is moving toward a state where the technical knowledge to build catastrophic weapons is becoming widely available. "Weapons-grade plutonium can be mail-ordered," he said, and many engineers now possess the skills to assemble nuclear devices—they just lack the designs.
AI changes the equation. Esvelt and other researchers worry that genius-level artificial intelligence could soon put advanced bioweapon capabilities within reach of actors with little scientific training—what some call the shift from "evil geniuses" to "evil morons" armed with AI.
U.S. policymakers now face pressure to act. The specific recommendations include tightening AI model access and regulations for labs that synthesize DNA sequences. Congress will need to weigh how aggressively to regulate the biotechnology sector without strangling legitimate research and innovation.
