The initiative by the UN Institute for Disarmament Research plans to launch on Thursday, establishing the first international forum for key players to try to tackle the issue.
Robin Geiss, head of the Geneva-based institute known as UNIDIR, warned at a UN press conference on Tuesday that "the margin for human judgment is increasingly narrow, including in the nuclear domain".
It's difficult to assess how close the world is to machines possibly acting outside of human control, Geiss said, and that uncertainty "is the issue we're trying to resolve here".
One problem is that AI is increasingly creeping into so many areas - early warning systems, information gathering and reconnaissance, to name a few - "and it's quite difficult to say at which point we're actually really losing control", he said.
The proliferation of AI capabilities and its potential impact have led to "a growing realisation that we stand at a crossroad - that now is the time to act, including on military AI".
Geiss said the UN recognises that global competition to lead on AI poses serious obstacles.
"But concerns about AI's impact are already transcending geopolitical blocs, and they're prompting action from major powers," he said, pointing to a proposal last month for the US and China to set up an AI safety alert system for issues that could threaten national security.
On Thursday, UNIDIR will launch a new forum as part of the centre to bring together "actors from every relevant sector and all over the world" to work on AI in military uses.
The UN agency will hold consultations this week with defence technology companies, software developers, subcontractors, investors and others that will work alongside governments, civil society, academia and experts, Geiss said.
UNIDIR also is launching an initiative focused on the risks of AI misuse by armed groups, he said, warning that such actors are already using the technology to support operational planning, execution and logistics.
"The urgent issue of AI in the military domain will certainly no longer be allowed to linger in the shadows," Geiss said.
"The time has come to pull our strengths and confront AI in the military head-on."
