Why the Department of War Has Issues With Effective Altruism

September 27, 2026 - 01:00

The global race for artificial intelligence is fueled by zero-sum thinking, probability neglect, and other cognitive biases, according to recent analysis. These mental shortcuts shape how governments and military institutions approach technology, often putting them at odds with the rational, long-term focus of effective altruism.

Effective altruism encourages people to use evidence and careful reasoning to determine how to do the most good. In the context of AI, that means prioritizing safety research, working to prevent catastrophic outcomes, and treating future generations as morally important. The Department of War, by contrast, operates on a different logic. Its mission centers on winning conflicts, deterring rivals, and maintaining strategic advantage. That framework naturally favors speed, secrecy, and dominance over caution and global coordination.

Zero-sum thinking assumes that one nation's gain must come at another's expense. In an AI arms race, this mindset pushes countries to cut corners on safety and dismiss cooperative solutions. Probability neglect leads decision-makers to ignore low-probability, high-impact risks, such as an AI system escaping human control. These biases make it hard for military planners to embrace effective altruism's call for restraint.

The tension is not just philosophical. It affects funding, policy, and research priorities. If the institutions driving AI development remain locked in adversarial thinking, the most serious risks may go unaddressed until it is too late.