170+ Dutch AI experts urge government to ban autonomous weapons

An open letter from more than 170 Dutch AI and robotics researchers, circulated as a signed document, landed at a specific moment in Dutch policy, and the timing is most of the story. The signatories, coordinated with peace NGO PAX and published via the University of Groningen's Campus Fryslân, are asking the Netherlands to back an international, legally binding agreement that preventively prohibits lethal autonomous weapons and locks in 'meaningful human control over the use of force.'

The framing they picked is worth paying attention to. Rather than argue about individual weapons systems, the letter leans on the credibility of the people signing it, professors from VU Amsterdam, TU Eindhoven, University of Amsterdam and Radboud plus international academics, and puts a category-wide argument on the table. Lambèr Royakkers, professor at TU Eindhoven, is quoted as saying it is 'simply inhumane to leave the killing of people to machines,' and the reporting flags examples like border-patrol robots that could kill without human intervention and drones that bomb vehicles based on license-plate recognition.

The timing is the important part. The letter was designed to land just as the Dutch government was preparing to review its position on autonomous weapon systems on the basis of an evaluation by the Advisory Council on International Affairs (AIV) and the Advisory Committee on Public International Law (CAVV). That is a narrow window where academic and NGO pressure can measurably shift where a small-but-influential arms-control voice ends up sitting. Similar national letters have been surfaced in Australia, Belgium, Canada, Italy and Norway, so the signatories are trying to add the Netherlands to that bloc.

The honest caveat is that a letter of this shape does not tell you what 'meaningful human control' actually looks like in code or in rules-of-engagement, and the retrieved reporting does not surface which specific programmes, Dutch defence contractors, or export lines would be affected. It is a directional ask, not a technical spec.

What is worth watching from here is whether the AIV/CAVV review, and the Dutch government's response, adopts the signatories' preventive-ban framing or lands on something softer around human oversight; that outcome, more than the letter itself, is what will set procurement and research constraints for any Dutch-tied AI vendor working near targeting.