Battles & Wars › Modern conflicts
AI and data-driven warfare
2017 onward (Project Maven); Ukraine and Gaza from 2022· Invention or discovery· Technology, codes & intelligence
Mostly agreedGood evidence; some details, dates or numbers uncertain.
Using machine learning to sift huge streams of sensor data, flag targets and steer weapons faster than human staffs can alone.
See it on the Battles & Wars map →What happened
Modern militaries collect more drone video, satellite images and intercepted signals than people can read. AI software sorts it: spotting vehicles, linking phone data to people, and proposing targets. The Pentagon's Project Maven began in 2017. Ukraine uses AI to fuse battlefield data and to guide drones past jamming. Israel's reported Lavender and Gospel systems drew scrutiny in Gaza. In the 2026 Iran war, Maven software reportedly helped identify about 1,000 targets in the first 24 hours.
Why it matters
It shrinks the 'kill chain', from spotting a target to striking it, from hours to minutes, raising hard questions about human judgment and accountability in war.
How we know
Pentagon memos and contracts, statements by US Central Command, Ukrainian officials and the Israeli military, investigative reporting by +972 Magazine and others, and UN debates on autonomous weapons since 2014.
Still debated
How closely humans review AI-suggested targets is disputed. Critics link fast AI targeting to errors such as the February 2026 strike on a school in Minab, Iran; officials say humans decide.
Inside this (4)
Connections (6)
▸Alan Turing asked 'Can machines think?', the question behind AI and data-driven warfareMostly agreed
The connection
Turing's 1950 paper 'Computing Machinery and Intelligence' asked whether machines could think and proposed a test. His wartime work had already shown machines could out-sift humans at finding patterns in enemy signals. Today's military AI, searching intercepts and imagery for targets, descends from both ideas.
When and how
Bletchley Park, 1939–45; the 1950 paper appeared in the journal Mind while Turing worked at Manchester University.
How sure are we?
His published papers and declassified Bletchley records; the link is intellectual lineage, since modern machine learning grew from much later research.
Debates and open questions
Turing did not foresee today's neural networks in detail, and what he would think of machines that help choose targets is unknowable.
▸AI and data-driven warfare is being tested at scale in the Russia–Ukraine warWell documented
The connection
Ukraine combines drone, satellite and soldier reports in software such as its Delta system and Palantir tools, using AI to spot targets and speed strikes. Both sides add AI to drones so they can lock onto targets when jamming cuts their links. Western companies test products there in real combat.
When and how
Across Ukraine's front since 2022, involving Ukrainian developers, Palantir and other Western firms, and Russian drone makers.
How we know
Statements by Ukrainian officials and companies, combat video, and reporting from the front; exact performance data remain classified.
Debates and open questions
How autonomous drones in Ukraine really are, and whether any already choose targets without human approval, is hard to verify.
▸AI and data-driven warfare drew scrutiny over target selection in the Israel–Hamas warDebated
The connection
In November 2023 and April 2024, +972 Magazine and Local Call, citing Israeli intelligence sources, reported that systems called the Gospel and Lavender generated tens of thousands of suggested targets, with little human review and permissive rules on civilian harm. Israel's military says these are aids for analysts and that humans make every decision.
When and how
Gaza, from October 2023; the reports appeared in November 2023 and April 2024, followed by debate at the UN and among AI experts.
How sure are we?
The claims rest on anonymous sources; Israel acknowledged using such tools but rejected the descriptions, so the extent of the link is debated.
Debates and open questions
How targets were really chosen and checked, and how much AI contributed to Gaza's high civilian toll, are disputed and may only emerge from future inquiries.
▸AI and data-driven warfare is giving growing autonomy to Drones and precision strikeMostly agreed
The connection
Most drones still need a human pilot and a radio link, which jamming can cut. AI lets drones recognize and lock onto targets in the final seconds, navigate without GPS, and fly in coordinated groups. Militaries are moving from remote-controlled aircraft toward weapons that decide more on their own.
When and how
Ukraine and Russia from about 2023; US programs such as the Pentagon's Replicator initiative (2023) and the Air Force's Collaborative Combat Aircraft.
How sure are we?
Manufacturers' specifications, US policy such as Defense Department Directive 3000.09 on autonomy in weapons, and combat footage.
Debates and open questions
Whether fully autonomous killing should be banned is debated at the UN, and how much autonomy deployed drones already have is often unclear.
▸Project Maven (Pentagon AI targeting) launched the Pentagon's push into AI and data-driven warfareWell documented
The connection
Maven was the Pentagon's first major program to put machine learning into combat operations. It grew from analyzing video to fusing many kinds of intelligence into target suggestions. Its successes and controversies set the pattern for later military AI, including the Pentagon's Chief Digital and AI Office, created in 2022.
When and how
Launched April 2017 under Deputy Defense Secretary Robert Work; later run by the National Geospatial-Intelligence Agency and Pentagon AI offices.
How we know
Official memos, budgets and contract records trace Maven's growth, and US commanders have described its use in operations.
Debates and open questions
Whether Maven speeds decisions without adding errors, and how it weighs the presence of civilians, cannot be judged from public information.
▸Palantir Technologies sells data and AI targeting software for AI and data-driven warfareWell documented
The connection
Palantir's software pulls together satellite images, drone video, radar and intercepted signals, then uses AI to flag likely targets and suggest weapons. It is used by the US military, NATO, Ukraine and Israel, making the company a main supplier of the data layer behind modern targeting.
When and how
Worldwide since the 2000s; in Ukraine from 2022, in a partnership with Israel's Defense Ministry from January 2024, and in the 2026 Iran war.
How we know
Contracts and partnership announcements, statements by CEO Alex Karp and Ukrainian officials, and US commanders' descriptions of Maven.
Debates and open questions
Critics fear private software firms gain too much influence over life-and-death decisions; how much its tools change outcomes is hard to verify.