Militaries globally are increasingly relying on artificial intelligence (AI) to find, track, and annihilate targets. This development has led to calls for a global policy and legal response, ideally through a treaty to regulate autonomous weapons systems.
An upcoming conference in Geneva, the seventh review conference of the Convention on Certain Conventional Weapons, offers an opportunity to address the risks associated with AI-enabled warfare. A key choice for states attending will be whether to launch negotiations on treaty provisions for these weapons.
AI systems can now facilitate strikes in minutes or even seconds, a significant reduction from the half-hour or more required for human consideration only a few years ago. This acceleration occurs when AI systems meet a confidence threshold set by a human operator.
Concerns have been raised regarding accountability when autonomous weapons systems are involved in potential war crimes, particularly with anti-personnel weapons designed to target people. These weapons may struggle to differentiate between combatants, journalists, civilians, or children.
Helen Clark, former head of the UN Development Programme and former prime minister of New Zealand, highlighted the widening gap between technological capability and regulatory constraint. She noted that governments with advanced autonomous weapons systems have resisted binding limits on their development and use.