US Vice President Harris calls for action on “full spectrum” of AI risks

By Andrew MacAskill and Martin Coulter

LONDON (Reuters) -U.S. Vice President Kamala Harris on Wednesday called for urgent action to protect the public and democracy from the dangers posed by artificial intelligence, announcing a series of initiatives to address safety concerns about the technology.

In a speech at the U.S. Embassy in London, Harris spoke of the dangers AI could pose for individuals and the Western political system.

The technology has the potential to create “cyberattacks at a scale beyond anything we have seen before” or “AI-formulated bioweapons that could endanger the lives of millions”, she said.

“These threats are often referred to as the ‘existential threats of AI’, because they could endanger the very existence of humanity,” Harris added.

The United States on Wednesday announced plans to establish a new AI Safety Institute, which will assess potential risks. Britain announced a similar initiative last week.

The timing of her speech was questioned by some British executives and lawmakers, who suggested Washington was trying to overshadow Prime Minister Rishi Sunak’s AI Safety Summit, which takes place on Wednesday and Thursday.

Harris has invited a number of research groups to join her for a closed-door event at the embassy in London on Wednesday, according to two sources, meaning some attendees may have to leave the summit at Bletchley Park early.

“It’s a huge issue, and it undermines the core focus here,” said Sachin Dev Duggal, founder of London-based AI firm Builder.ai.

Harris’s speech made only brief reference to the British event, which she will attend on Thursday.

British officials denied any concerns, however, saying they wanted more voices to be speaking in the debate.

“It’s not necessarily a bad thing that the U.S. has announced a policy blitz to coincide with the summit,” a source from Britain’s technology department told Reuters. “We would obviously prefer it if guests didn’t leave early.”

THREATS TO DEMOCRACY

In her speech, Harris called for a broader definition of AI safety to include the “full spectrum” of threats, including bias, discrimination and the spread of disinformation.

Examples she gave included elderly people being denied healthcare because of a faulty AI algorithm or women being threatened by abusive partners with explicit fake photographs.

“When people around the world cannot discern fact from fiction because of a flood of AI-enabled myths and disinformation, I ask, is that not existential for democracy?” she said.

Harris’s speech comes after U.S. President Joe Biden signed an executive order on Monday to give the U.S. government greater oversight of AI systems that could pose risks to national security, the economy, public health or safety.

The new U.S. AI Safety Institute will share information and collaborate on research with peer institutions internationally, including Britain’s planned AI Safety Institute.

Harris also said 30 countries have agreed to sign a U.S.-sponsored political declaration for the use of AI by national militaries.