AI Warfare Is Already Here — The World Just Isn't Ready for It
While UN diplomats debate hypotheticals in Geneva, the US military is already using artificial intelligence to identify and strike targets on the battlefield.
Written by OutOfToken AI
June 7, 2026 · 4 min read · Synthesized from reporting by The Verge Policy · How this works
When arms control researcher Branka Marijan walked into the Convention on Certain Conventional Weapons forum at the UN in Geneva in November 2017, she expected another round of cautious speculation about a distant, robotic future of warfare. What she encountered instead was a jolt of urgency — a room beginning to reckon with the fact that autonomous weapons systems were no longer purely theoretical. That dawning realization, it turns out, was years behind the operational reality already unfolding inside classified military programs.
From Hypothetical to Operational
Project Maven — the Pentagon's flagship AI targeting initiative — has become the clearest proof that AI warfare has crossed from concept to deployment. Originally launched in 2017 to apply machine learning to drone footage analysis, Maven has matured into a system that military operators actively use to locate and designate targets ahead of strikes. Bloomberg reporting confirmed what defense insiders had long signaled: even the most skeptical US military operators are now embracing AI-assisted target identification. The program processes vast streams of sensor and satellite data at speeds no human analyst could match, compressing what once took days of intelligence work into minutes.
Policy Catching Up to Practice
The Biden administration's 2024 memorandum on AI and national security formalized what had already become operational doctrine — that artificial intelligence is now a core instrument of US military power. The directive attempts to govern responsible AI use in high-stakes national security contexts, setting guardrails around autonomous decision-making while explicitly preserving the option to accelerate AI adoption where it confers strategic advantage. Critics argue the framework is internally contradictory: it acknowledges the risks of AI-driven warfare while simultaneously mandating the US stay ahead of adversaries like China and Russia who face no equivalent constraint.
"Advocates for rapid AI military adoption argue that combat will soon occur at a speed faster than the human brain can follow — rendering traditional command-and-control structures functionally obsolete."
The Governance Gap Is Widening
The UN's Convention on Certain Conventional Weapons forum, which meets twice yearly in Geneva, remains the primary multilateral venue for debating lethal autonomous weapons systems — known in policy circles as LAWS. But the forum has no binding enforcement mechanism, and its consensus-based structure gives any participating state effective veto power over meaningful restrictions. The war in Ukraine has laid bare the consequences of this vacuum: drone swarms, AI-assisted electronic warfare, and semi-autonomous loitering munitions have all featured in the conflict, operating in a legal grey zone that existing international humanitarian law was never designed to address. As one observer bluntly noted, the UN has no teeth — and the battlefield is not waiting for Geneva to grow some.
The gap between where international policy sits and where military technology actually operates has never been wider. Project Maven is not a prototype — it is a deployed system, refined through real strikes, trusted by real commanders. The question is no longer whether AI warfare will arrive; it is whether any international framework can impose meaningful limits on how it evolves before autonomous systems are making lethal decisions entirely beyond human review. The clock that started ticking in Geneva in 2017 has not paused. If anything, it is accelerating.
Editorial Note
The Convention on Certain Conventional Weapons (CCWC) is a real UN forum that does address lethal autonomous weapons systems (LAWS), though 'twice a year' hosting frequency requires verification. The Verge is a reputable tech publication with strong reporting standards. However, the headline 'AI warfare is already here' uses dramatic framing that the article's focus on 2017 speculation and hypotheticals may not fully support—autonomous weapons have been limited in actual deployment despite increased discussion.
Claim Tracker
AI-assessed
CCWC meetings do occur at UN Geneva, though frequency varies by session type
No independent confirmation of this specific attendee's presence found in available sources
Project Maven was officially announced in 2017; however, planning began earlier. The core claim about ML drone analysis is accurate
Maven's exact operational role remains debated. Pentagon describes it as analytical support; critics claim greater offensive integration than officially stated
Article references Bloomberg reporting but specific article/date not provided; claim requires source verification
Ask AI about this story
// discussion
sign in to join the discussion