Time is running out for regulators to address the potential threats posed by artificial intelligence (AI) to the democratic process before the upcoming July election, according to a new report published by the Alan Turing Institute thinktank.
While there is currently no clear evidence of AI directly affecting election results, the report’s authors caution against underestimating the impact of AI-generated content on the democratic system. In recent months, there has been a surge in AI-generated material depicting political figures, with examples such as deepfake audio of London mayor Sadiq Khan and US President Joe Biden.
The lead author of the report, Sam Stockwell, emphasizes the potential consequences of this type of technology, stating, “The challenge of differentiating between AI-generated and authentic content poses numerous issues. It allows bad actors to exploit this uncertainty and dismiss deepfake content as mere allegations, undermining the fundamental principles of democracy.”
To combat this growing threat, the report calls for joint guidance from Ofcom and the Electoral Commission, as well as voluntary agreements from political parties on the fair use of AI in election campaigning. It also recommends that media outlets receive guidance on reporting about AI-generated content and that voter information includes instructions on how to identify and seek advice about AI-generated content.
Dr. Christian Schroeder de Witt, a researcher at the University of Oxford, suggests several potential solutions, including adding watermarks to AI-generated content and implementing provenance-based solutions that tie the image and metadata together to detect any tampering.
With the general election only weeks away, Mr. Stockwell stresses the urgency for regulators to take action before it is too late. “There is currently no clear guidance or expectations for preventing the use of AI to create false or misleading electoral information,” he states.
Dr. Alexander Babuta, director of the Centre for Emerging Technology and Security at the Alan Turing Institute, acknowledges that there is no evidence of AI directly impacting election results, but urges action to ensure the resilience of future elections. “Regulators can do more to help the public distinguish fact from fiction and ensure voters don’t lose faith in the democratic process,” he says.
In response to the report, a spokesperson for the Electoral Commission states, “We are well-equipped to combat false information about the voting process and the administration of elections.” They also mention plans to provide factual and accurate information to voters and to address misleading information about election procedures.
Similarly, a spokesperson for Ofcom states that the organization will carefully consider the report’s recommendations. However, they note the challenges of regulating campaign material and the need for a new legal framework.
As the world’s biggest election year approaches, it is crucial for regulators to take swift action to address the potential threats of AI to the democratic process. The integrity of elections and the trust of voters are at stake, and it is essential to act now to prevent any potential damage.
