More than 60 countries including the US and China signed a modest “call to action” endorsing the responsible use of artificial intelligence (AI) in the military.
Human rights experts and academics noted the statement was not legally binding and failed to address concerns like AI-guided drones, ‘slaughterbots’ that could kill with no human intervention, or the risk that an AI could escalate a military conflict.
However, the statement was a tangible outcome of the first international summit on military AI, co-hosted by the Netherlands and South Korea this week at The Hague.
Signatories said they were committed to developing and using military AI in accordance with “international legal obligations and in a way that does not undermine international security, stability and accountability.”
The conference comes as interest in AI is at all-time highs thanks to the launch of OpenAI’s ChatGPT program.
Israel participated in the conference but did not sign the statement.
US under secretary of state for arms control Bonnie Jenkins put forward a US framework for responsible military AI use.
The US and other powerful countries have been reluctant to agree to any legal limitations on using AI, for fear that doing so might put them at a disadvantage to rivals.
“We want to emphasize that we are open to engagement with any country that is interested in joining us,” Jenkins said.
The US proposal said AI-weapons systems should involve “appropriate levels of human judgment”, in line with updated guidelines on lethal autonomous weapons issued by the Department of Defense last month.
Human Rights Watch challenged the US to define “appropriate”, and not to “tinker with political declarations” but to begin negotiating internationally binding law.
China representative Jian Tan told the summit that countries should “oppose seeking absolute military advantage and hegemony through AI” and work through the United Nations.
Jessica Dorsey, assistant professor of international law at Utrecht University, said the US proposal was a “missed opportunity” for leadership and the summit statement was too weak.
“It paves the path for states to develop AI for military purposes in any way they see fit as long as they can say it is ‘responsible’,” she said.
“Where is the enforcement mechanism?”