AI could 'supercharge' election disinformation, US tells the BBC

Artificial intelligence (AI) threatens to "supercharge" disinformation and incite violence at elections, the US deputy attorney general has warned.

Speaking exclusively to the BBC, Lisa Monaco described AI as the "ultimate double-edged sword".

It could deliver "profound benefits" to society but also be used by "malicious actors" to "sow chaos", she added.

And she revealed plans to make the use of AI by criminals an aggravating factor in sentencing in US courts.

The former federal prosecutor, who is in the UK to deliver a lecture on AI at the University of Oxford, said violent criminals who used guns were given longer sentences.

"So we are going to be applying that same principle and seeking stiffer sentences and sentencing enhancements for those that use AI in a malicious way to commit their crime," she told the BBC.

• Should we fear an attack of the voice clones?

• Fears UK not ready for deepfake general election

• Fake audio of Sadiq Khan is not a crime, says Met

She also spoke about efforts to protect the US election in November from being disrupted by AI-powered misinformation such as deepfakes - convincing audio and video of politicians saying things they never said - and AI-generated fake calls, or robocalls.

Last week, the US Federal Communications Commission (FCC) made robocalls during elections illegal.

It came after thousands of voters in New Hampshire received a phone call claiming to be from US President Joe Biden urging them not to vote in January's primary election.

Ms Monaco said: "That is the type of action you are going to see regulators, I think, appropriately taking to try and put some guardrails around the use of these technologies, particularly when it comes to the election space."

She said the US government was working with tech companies and other nations, including the UK, to combat the threat posed by AI.

But she added: "I do think we are going to see more of this, I think it is something we are just beginning to see the surface being scratched about how malicious actors can use this technology."

With more than two billion people eligible to take part in elections around the world this year, including in the US, UK and India, Ms Monaco said she was concerned AI could have a "fundamental impact" on democracy.

This week, London mayor Sadiq Khan told the BBC deepfake audio of him supposedly making inflammatory remarks before Armistice Day almost caused "serious disorder".

And the US Deputy Attorney General said she too was concerned about violence resulting from AI-powered misinformation.

"I am absolutely worried... about this effort by malicious actors, nation states or otherwise, using AI generated content to spread and really supercharge mis and disinformation.

"So that can have a number of effects. It can cause people to distrust the sources of information they are getting, to dissuade them or confuse them in terms of exercising their right to vote. To incite violence, certainly that's something that we are worried, about and to just generally sow distrust and potentially chaos."

Ms Monaco was also keen to stress the potential benefits of AI as a crimefighting tool and something that could help in investigations and prosecutions.

The FBI was using AI technology to "sift through the tips that we get from the public" and to analyse data and images in "some of our most important investigations" including the US Capitol riots on 6 January 2021, she told the BBC.

Ultimately, she said, it would require a combination of action by tech firms and legislation to set the appropriate "guardrails" for AI and protect democracy.