Iran war marks first AI-driven conflict as 2,000 strikes executed in 48 hours
Palantir CTO's claim that Iran-Israel hostilities represent a watershed in autonomous targeting gains credibility as operational tempo outpaces governance frameworks.
Palantir Technologies CTO Shyam Sankar stated that the Iran-Israel conflict will likely be remembered as the first major war in which artificial intelligence played a central operational role, a claim substantiated by the execution of 2,000 strikes in 48 hours using AI-enabled planning systems.
The assessment, Bloomberg reported on 24 March, positions the conflict as a tactical proof-of-concept for AI-accelerated targeting and a strategic inflection point comparable to WWI’s industrial warfare transition. The US military’s deployment of Anthropic’s Claude large language model through Palantir’s Maven Smart System collapsed decision cycles from days to seconds, per Bloomsbury Intelligence and Security Institute analysis of the campaign launched 28 February.
Operational validation through speed asymmetry
The strike tempo represents a quantitative break from prior conflicts. “The planning that happened was done in a fraction of the time it would have taken in prior conflicts of this scale, and we accomplished more than twice as much per day of strikes,” Sankar told Fox News. Admiral Brad Cooper, CENTCOM commander, confirmed to Al Jazeera that AI systems “turn processes that used to take hours and sometimes even days into seconds.”
The technological architecture fuses satellite imagery, signals intelligence, and predictive analytics at machine speed. Palantir’s Maven system, integrated with Claude, handled intelligence assessments, target identification, and battle simulations across the campaign’s opening phase. Israel simultaneously deployed its Habsora targeting system—previously tested in Gaza with a 20-second approval cycle and 10% accepted error rate—creating parallel AI-enabled strike channels, per Common Dreams reporting.
Iran’s response underscored the asymmetry. By 5 March, Iranian forces had launched 500+ ballistic missiles and approximately 2,000 drones against US and Israeli targets, New Space Economy documented. The volume-based strategy contrasted sharply with precision-targeted strikes processed through AI fusion systems—a divergence that establishes competing models for future conflicts.
“Our war fighters are leveraging a variety of advanced AI tools. These systems help us sift through vast amounts of data in seconds so our leaders can cut through the noise and make smarter decisions faster than the enemy can react.”
— Admiral Brad Cooper, Commander, US Central Command
Civilian casualties expose governance vacuum
A 1 March strike on Shajarah Tayyebeh girls’ elementary school in southern Iran killed 175 people, predominantly children. The Pentagon opened an investigation into whether AI targeting systems played a role in selecting or approving the school as a valid target, Foreign Policy reported. The incident crystallises the operational risk Craig Jones, a warfare expert at Newcastle University, outlined: “There is no evidence that AI lowers civilian deaths or wrongful targeting decisions—and it may be that the opposite is true.”
The strike occurred within the operational window when Maven and Claude were processing thousands of targets through accelerated approval chains. Jones noted to Democracy Now that such systems “reduce a massive human workload of tens of thousands of hours into seconds and minutes,” compressing decision cycles but simultaneously automating choices that “open up all kinds of problematic legal, ethical and political questions.”
The Pentagon terminated Anthropic’s Maven contract after the company imposed restrictions on unrestricted military use of Claude. OpenAI replaced Anthropic, promising deployment guardrails but providing no operational evidence of their implementation, Foreign Policy noted. The substitution occurred mid-campaign, creating continuity questions around human oversight protocols.
International law framework fails real-time adaptation
The conflict exposed the inadequacy of existing international humanitarian law to govern Autonomous Weapons deployment. The International Committee of the Red Cross stated in a 2 March assessment that current IHL “does not hold all the answers” on autonomous weapon systems, calling for new rules tailored to AI capabilities.
The UN has set a 2026 deadline for completing a legally binding treaty on lethal autonomous weapons systems, with 156 nations backing resolution language for negotiations, according to the Usanas Foundation. The US and Russia opposed the resolution. Israel abstained, citing national security requirements for autonomous defensive systems in analysis by the West Point Lieber Institute.
The Iran campaign occurred precisely as diplomatic efforts approached their deadline, creating operational precedent before regulatory frameworks could solidify. Paul Scharre, executive vice president at the Center for a New American Security, told The Week that “ours is now officially an ‘age of AI warfare'”—a characterisation that acknowledges the regulatory gap has closed through battlefield adoption rather than treaty completion.
Deterrence model under revision
The operational tempo forces immediate recalibration of force planning assumptions. Sankar framed human-AI collaboration as “Luke Skywalker and R2-D2″—a team model of “human computer symbiosis” rather than sequential approval chains. This architecture implies persistent AI presence in targeting loops rather than episodic consultation, fundamentally altering the threshold between human oversight and autonomous action.
The US military deployed autonomous drone swarms with embedded AI flight controls during the campaign, New Space Economy reported. These LUCAS systems operated with onboard decision-making capabilities, extending the autonomy spectrum beyond targeting assistance into execution domains. The dual deployment—AI-enabled planning through Maven alongside autonomous platforms—establishes layered automation that compounds governance challenges.
- Decision cycle compression creates first-strike advantages that destabilise deterrence equilibria built on warning time assumptions
- Civilian casualty investigations will test whether IHL concepts of proportionality and distinction retain operational meaning at machine speed
- AI system proliferation to peer adversaries becomes inevitable absent enforceable export controls—China and Russia will reverse-engineer operational concepts within 18-24 months
- Human-in-loop doctrine requires redefinition: oversight measured in seconds differs categorically from deliberative approval
What to watch
The Pentagon’s investigation into the Shajarah Tayyebeh school strike will establish precedent for attributing responsibility when AI systems participate in target selection. Results are expected before mid-2026 and will indicate whether existing accountability frameworks can accommodate algorithmic decision-making or require legislative intervention.
UN treaty negotiations face a compressed timeline. Diplomatic progress toward the 2026 deadline will reveal whether major powers accept binding constraints or pursue unilateral capability development. Current US and Russian opposition suggests the latter trajectory, with treaty language likely to codify lowest-common-denominator standards that permit continued autonomous weapons expansion.
Defence contractors beyond Palantir will accelerate AI integration offerings. Anduril, Shield AI, and defense primes with AI divisions will cite Iran operational tempo as validation, driving procurement decisions across NATO and Pacific allies. Budget allocations in the 2027 US defense appropriations cycle will quantify institutional commitment to AI-enabled systems.
Adversary adaptation timelines present the critical variable. If China demonstrates comparable AI targeting fusion within 12 months, the brief US-Israeli advantage collapses into peer competition with higher escalation risk. Iranian lessons-learned assessments—particularly on countering AI-enabled targeting through deception, dispersion, and saturation—will shape the next phase of conflict escalation.