Military AI Coalition Of 13 Countries Meets On Ethics

WASHINGTON: In an extraordinary meeting that highlights how crucial artificial intelligence is becoming to the US and its allies, some 100 officials from 13 democratic countries met online Tuesday and Wednesday to discuss how their militaries could ethically use AI, the first summit of its kind. Hosted by the Pentagon’s Joint AI Center, the virtual conference kicked off what JAIC’s calling the AI Partnership for Defense, an international forum it hopes will evolve from broad principles and policy to technical cooperation on data and algorithms.

“This is historic,” said Mark Beall, who’s been working on international cooperation since he became the JAIC’s chief strategy and policy 18 months ago. “This group of my countries, to my knowledge, has never been brought together under one banner before.”

Defense Secretary Mark Esper officially adopted a set of AI ethics principles in February, although implementation is still nascent. He has also repeatedly denounced China and Russia for developing, deploying, and in some cases exporting AI systems that disrespected human rights or human control of lethal force.

But this week’s conference is not about creating a US-led bloc in opposition to any nation, Beall says.

“We want to be inclusive as we can,” Beall told me. “It’s not about countering anybody. It’s really about a coming together of like-minded countries. If we saw… our competitors adopt responsible AI technology, we’d welcome that,” he said. That said, “we’re really focused on, right now, rallying around [shared] core values like digital liberty and human rights… international humanitarian law.”

What came together this week was a broad coalition of the willing, based on which countries were most enthusiastic about JAIC’s outreach – a dozen formal and informal US allies:

• NATO members: Britain, Canada, Denmark, Estonia, France, and Norway.
• Non-NATO treaty allies: Australia, Japan, and South Korea.
• De facto allies: Israel – which cooperates closely with the US on cyber warfare, intelligence, and arms, despite the lack of a formal mutual-defense pledge – and Finland and Sweden, Cold War neutrals who remain outside NATO but are increasingly aligned with it in practice.

About 20 US personnel participated; the other 80-odd were foreign delegates.

While Beall doesn’t see China or Russia joining the Partnership for Defense any time soon, he hopes it will expand to more of the “40 to 50” friendly nations JAIC has been in touch with.

This week’s event began with a discussion of ethical principles and military use cases for AI on Tuesday, then delved into implementation on Wednesday. JAIC’s aiming for three such meetings a year.

“We’re starting with policy challenges and then hopefully over time more military technical cooperation can develop,” Beall told me. “My personal goal for this forum is to create a framework for data sharing and data aggregation [and collaboration on] very powerful, detailed algorithms.”