The US and Russia support eliminating certain rules concerning human control and the predictability of lethal autonomous weapons

The United States and Russia supported removing several safeguards from the UN draft on lethal autonomous weapons during negotiations held in Geneva at the beginning of September. The changes concern human control, system predictability, and compliance with ethical criteria in the use of artificial intelligence on the battlefield.

After approximately 15 hours of discussions on the final day, the requirement that a person verify targets identified by autonomous systems before an attack was also eliminated. The document no longer includes the requirement that these systems operate in a predictable and reliable manner.

The negotiations, held behind closed doors, sought to develop the first legally binding international agreement on autonomous weapons. According to The Washington Post, the US and Russian delegations each had approximately ten lawyers, while the rapidly introduced changes reportedly made it more difficult for states with smaller delegations to respond.

The move comes amid the Pentagon’s expanding use of artificial intelligence and the revision of US rules on autonomous weapons. The 2023 directive required the exercise of human judgment and compliance with international humanitarian law. Human Rights Watch warns that weakening the safeguards could allow life-and-death decisions to be made with insufficient human control and could reduce accountability in the event of attacks.