Topic
The tendency of human operators to defer to machine judgments over their own, and how AI may intensify it.
Officials told Bloomberg some Centcom staff relied too much on Palantir's Maven AI before a US strike killed more than 165 at an Iranian girls' school. The reported failure is users trusting a fused screen too far. Builders of AI decision support have to design for that risk.
Reality
Researchers found an auto-displayed AI answer cut 'I don't know' responses from 35 percent to 1 percent, though the model was mostly wrong. A 10-cent penalty for wrong answers still left abstention at 7 percent, so review tools need more than an abstain button.
Reality
A Bloomberg investigation into an unreleased Pentagon review says CENTCOM personnel expected the Maven Smart System to flag stale or contradictory intelligence by itself. Palantir says the underlying data was never its responsibility.
Reality
Wharton's Scale of Agency Decay runs experimenting, integrating, relying, depending. The account details the first two, and its usable marker is whether people still overrule a system that contradicts them.
Reality
In a Harvard, MIT and University of Washington experiment, 228 reviewers deferred to an LLM at the same rate with or without a rationale. Only the silent version improved their accuracy.
Reality
ECRI put AI in clinical diagnosis at number one on 9 March 2026. The absence of hearings, moratoria, or named vendors tells you what shape the constraint will actually take.
Reality
A Foreign Affairs essay argues the real military AI risk is cognitive: automation bias and deskilling hollowing out the judgment a chain of command assumes it has. Delegation is the exposure.
Reality