The Defense Innovation Board, a Pentagon advisory board led by Eric Schmidt, unveils principles for the US military to apply to combat and non-combat AI systems
Khari Johnson / VentureBeat :
Context & Ripple Effects
This is the advisory layer of a Pentagon AI build-out already in motion: the Algorithmic Warfare Cross Functional Team was created for drone-video analysis in 2017, and the Department followed by planning $1.7B over five years for the Joint Artificial Intelligence Center. With Eric Schmidt chairing the board, the principles are an attempt to give that spending an ethical frame covering both combat and non-combat systems.
The arc matters because these recommendations were not the endpoint — the DOD formally adopted its own set of AI ethics principles months later after 15 months of consultation with leading AI experts, and years on the doctrine has shifted further toward machine initiative. Schmidt himself is not a neutral observer: he backs defense startups including Ishtari and Rebellion Defense, which builds AI applications for the US government.
First-order effects
- DOD acquisition and program offices now have a written framework they can be held to when fielding combat and non-combat AI, giving the JAIC and its contractors a common compliance baseline.
- Schmidt gains formal standing at the intersection of his advisory role and his investments in government-facing defense AI firms like Rebellion Defense and Ishtari.
Second-order effects
- Defense AI vendors must build their offerings around the principles' requirements — human judgment over lethal force, auditability — turning ethics compliance into a competitive differentiator for contracts.
- Rival advisory bodies and services will push their own interpretations, forcing the Pentagon to reconcile board recommendations with operational doctrine rather than treating them as parallel documents.
Third-order effects
- If the pattern holds, advisory ethics frameworks get codified into official policy but then bend toward operations: the later revision of targeting doctrine envisioning "systems where AI initiates actions with human monitoring" shows the guardrails loosening as capability matures — whether oversight keeps pace is the open structural question.
- The episode establishes a template other militaries copy: convene a tech-industry-led advisory board first, adopt its principles as legitimacy cover, then scale deployment under them — making civilian tech leaders permanent fixtures of defense AI governance.
The trend: Military AI governance is evolving from industry-advised ethics codes into operational doctrine that steadily enlarges AI's role in decisions about force.