As worries about the impending next-generation versions of artificial intelligence mount, the United States and Britain established on Monday a new alliance in the field of artificial intelligence safety science.
Following promises made at an AI Safety Summit in November at Bletchley Park, Secretary of Commerce Gina Raimondo and Secretary of Technology for Great Britain Michelle Donelan signed a memorandum of understanding in Washington to work together to create improved AI model testing.
“We all know AI is the defining technology of our generation,” stated Raimondo. “This partnership will accelerate both of our institutes work across the full spectrum to address the risks of our national security concerns and the concerns of our broader society.”
Among the nations creating government-run AI safety institutions are the United States and Britain.
While the United States announced in November that it was establishing its own safety institute to assess the hazards associated with so-called frontier AI models and is currently collaborating with 200 organizations and entities, Britain indicated in October that its institution would investigate and test new forms of AI.
As part of their official collaboration, the United States and Britain intend to do joint testing on a model that is accessible to the public at least once, and they are also thinking about looking into staff exchanges between the institutes. Both are attempting to establish comparable alliances with foreign nations in order to advance AI safety.
This is the world’s first agreement of its kind, according to Donelan. “AI is already an extraordinary force for good in our society, and has vast potential to tackle some of the world’s biggest challenges, but only if we are able to grip those risks.”
With the ability to produce text, images, and videos in response to open-ended cues, generative AI has sparked both enthusiasm and concerns that it would eventually replace human labor in some occupations, disrupt elections, and have disastrous consequences.
Together, Raimondo and Donelan told Reuters on Monday that immediate cooperation was required to handle the threats associated with AI.
Since the next generation of models—which will be far, far more capable—will soon be produced, time is of the importance, according to Donelan. “We have a focus one the areas that we are dividing and conquering and really specializing.”
Raimondo announced that she would bring up AI-related topics during the U.S.-EU Trade and Technology Council meeting on Thursday in Belgium.
According to Raimondo, the Biden administration will shortly reveal new members of its AI team. “We are pulling in the full resources of the U.S. government.”
Both nations intend to exchange important data on the risks and potential benefits of AI models and systems, as well as technical research on AI security and safety.
An executive order to lessen the hazards associated with AI was signed by Biden in October. The Commerce Department announced in January that it would like to make American cloud providers check if foreign organizations are using American data centers to train artificial intelligence (AI) models.
In February, the United Kingdom announced that it would invest over 100 million pounds ($125.5 million) to establish nine new research centers and educate regulators on artificial intelligence.
Raimondo expressed particular concern about the potential for AI to be used in nuclear war simulations or bioterrorism.
“Those are the things where the consequences could be catastrophic and so we really have to have zero tolerance for some of these models being used for that capability,” she stated.
