The Associated Press spoke about the dangers of artificial intelligence tools for elections, starting in 2024, especially after they have become accessible to everyone, due to their cheapness and power.
And the agency said in a report, Sunday, that computer engineers, political scientists and tech geeks warned that artificial intelligence tools will soon allow anyone to fake images, videos and sounds that are realistic enough to deceive voters and possibly influence elections.
The agency stated that unlike synthetic images that appeared suddenly and are considered unconvincing and expensive to produce, the threat posed by artificial intelligence and so-called deep fakes always seemed a year or two away.
But now, artificial intelligence tools can create cloned human voices, ultra-realistic images, videos and audio in seconds and at the lowest cost, according to the agency.
This digitally generated fake content with powerful algorithms, when linked to social media, can easily spread widely and quickly and target very specific audiences, which may promote fraud and dishonest campaign goals.
According to the agency, the repercussions of artificial intelligence on the 2024 campaigns and elections will be as great as they are worrisome, as generative artificial intelligence can not only quickly produce emails, texts, or videos, but it can also be used to mislead voters, impersonate candidates, and undermine elections. Widely.
And the "Associated Press" indicated that artificial intelligence experts expected a number of disturbing scenarios in which synthetic artificial intelligence is used to create synthetic media for the purposes of confounding voters, defaming a candidate, or even inciting violence.
Such scenarios include sending automated messages with the candidate's voice to instruct voters to cast their ballots on the wrong date, the release of audio recordings of a candidate supposedly confessing to a crime or expressing racist views, or the fabrication of video footage showing a person giving a speech or interview that never took place. It is also possible to easily create fake images designed to look like local news reports falsely claiming a candidate has withdrawn from the race.
"What if Elon Musk called you personally and asked you to vote for a particular candidate?" Oren Etzioni, the founding CEO of the Allen Institute for Artificial Intelligence, who quit last year to start the AI2 nonprofit, told the agency. "A lot of people will listen. But in reality it's not him," he added.
The agency quoted the global chief technology officer of Forcepoint, a cybersecurity company based in Austin, Texas, as saying that the US elections in 2024 could be manipulated by nefarious political campaigns and foreign adversaries due to artificial intelligence.
Stoyanov predicted that groups looking to interfere with American democracy will use artificial intelligence and synthetic media as a way to undermine trust. "We will see a lot of misinformation from international sources," Stoyanov said.
To combat the threat of artificial intelligence to the elections, the agency stated that the New York Democrat, Yvette Clark, introduced legislation in the US House of Representatives, obligating candidates to name campaign ads that were created using artificial intelligence, and the legislation also requires anyone who creates artificial images to add a watermark indicating the truth.
According to the agency, some countries have put forward their own proposals to address concerns about AI deepfakes.