US, UK Form Historic Alliance on AI Safety, Testing

The U.S. and U.K. have pledged to jointly develop tests to evaluate the safety of AI systems in a first-of-its-kind bilateral agreement

U.S. Commerce Secretary Gina Raimondo and U.K. Technology Secretary Michelle Donelan signed a Memorandum of Understanding to facilitate the exchange of information and technical knowledge regarding AI model risks and associated security measures between both nations.

The announcement follows the launch of new requirements for U.S. federal agencies on the responsible use of AI.

Under the agreement, both nations’ dedicated AI Safety Institutes will work together to create safety testing measures.

“This partnership is going to accelerate both of our Institutes’ work across the full spectrum of risks, whether to our national security or to our broader society,” Raimondo said. “Our partnership makes clear that we aren’t running away from these concerns – we’re running at them. Because of our collaboration, our Institutes will gain a better understanding of AI systems, conduct more robust evaluations and issue more rigorous guidance.”

Both the U.S. and U.K. founded AI safety institutes in the wake of last year’s AI Safety Summit. The agencies are tasked with creating respective guidelines and benchmarks for evaluating AI risks.

Related:White House Unveils Guidelines for Safe, Secure, Responsible AI Use

Researchers at the U.S. AI Safety Institute (U.S. AISI) will share capabilities with its U.K. counterpart to build a “common approach to AI safety testing.”

It was announced that joint testing exercises will also occur on publicly accessible models, although no specific time frame was provided for when these will take place.

U.K. Tech Secretary Michelle Donelan said the agreement represents a landmark model deepening the relationship between the two nations to address “the defining technology challenge of our generation.”

“We have always been clear that ensuring the safe development of AI is a shared global issue,” Donelan said. “Only by working together can we address the technology’s risks head-on and harness its enormous potential to help us all live easier and healthier lives.”

The tests are a result of the U.S.’ AI partnership with the U.K. and could prove helpful to startups, according to Ekaterina Almasque, general partner at venture capital firm OpenOcean.

“Startups in AI often encounter difficulties navigating the complex landscape of safety and ethics, such as sourcing ethical training data at a reasonable cost, which can impede their ability to drive innovation, scale and create competitive products,” Almasque said. “This collaboration provides a framework for addressing these challenges, offering guidance and support to help startups develop AI technologies responsibly and securely.”

Related:UK's AI Safety Summit: What They Discussed