The United Nations and the International Committee of the Red Cross (ICRC) have renewed pressure on governments to outlaw autonomous weapons that can make lethal decisions without meaningful human control, warning that rapid advances in artificial intelligence are bringing the world closer to a point where machines could independently decide who lives or dies.
UN Secretary-General António Guterres and ICRC President Mirjana Spoljaric Egger made the appeal ahead of the seventh Review Conference of the Convention on Certain Conventional Weapons (CCW), scheduled for November in Geneva.
“We are now dangerously close to crossing a moral red line: the autonomous targeting of humans by machines,” Guterres said in a joint appeal with Spoljaric Egger.
The two organisations want governments to move beyond years of discussions and begin negotiations on a legally binding agreement that would set clear bans and restrictions on autonomous weapons.
The call comes as weapons systems become more advanced and are developed and deployed at greater speed and scale, with increasingly powerful AI technology helping drive their capabilities.
The UN and ICRC warned that such systems could be left to make life-and-death decisions on fast-changing battlefields, where distinguishing between civilians and combatants can be difficult.
International humanitarian law already prohibits attacks on civilians and civilian infrastructure. However, the organisations said identifying people who are not taking part in fighting can become particularly difficult when conditions on the battlefield change within moments.
ICRC arms expert Laurent Gisel pointed to situations in which a combatant could surrender or become wounded and therefore no longer be a lawful target, while civilians could move into an area where autonomous weapons are operating.
“The weapons may encounter a situation that would not have been anticipated,” he said.
Gisel said the risks posed by modern warfare cannot be viewed only through the lens of precision or restraint. He pointed to civilians being killed and families forced from their homes, alongside the destruction of hospitals, schools and houses.
The joint appeal calls for two types of autonomous weapons to be banned.
The first covers unpredictable autonomous weapons whose operators cannot predict or anticipate the outcome of the force they use.
“If it's not possible to predict it, then how is it possible to limit it?” Gisel asked.
The second category involves anti-personnel autonomous weapons designed to target humans. The ICRC said these systems raise particular concerns because they could reduce human life to “data, algorithm, sensor and machine operations”.
The UN and ICRC made a similar joint appeal in 2023, urging States to negotiate a legally binding agreement by 2026.
With the November conference approaching, the organisations are now calling on governments to use the meeting as an opportunity to begin the negotiations.
They warned that delaying action could make it increasingly difficult to determine who should be held responsible when autonomous systems make lethal decisions.
“Innovation must serve humanity – not endanger it,” the leaders said.