AI Threats Loom Over Scottish Parliament Election

As Scotland braces for its parliamentary elections on May 7, election officials are preparing to combat an anticipated surge in artificial intelligence-generated content designed to manipulate voters. Experts warn that these sophisticated technologies pose significant risks to democratic processes not just in Scotland, but worldwide.

Bot networks, deepfakes, and AI chatbots are emerging as powerful tools that could potentially influence electoral outcomes through coordinated misinformation campaigns. These threats have evolved rapidly, becoming increasingly difficult to detect and counter.

Last month, a global consortium of experts, including Nobel Peace Prize winner Maria Ressa and researchers from prestigious institutions like Cambridge and Oxford, issued a stark warning about “swarms of collaborative, malicious AI agents.” These systems, they cautioned, can “coordinate autonomously, infiltrate communities and fabricate consensus efficiently,” directly threatening democratic processes by mimicking human social dynamics.

The creation of fake social media accounts—known as bot farming—has become increasingly sophisticated and affordable. Recent analysis from Cambridge University found that verifying fake accounts for use in the UK costs nearly the same as in Russia, while prices typically spike in countries approaching national elections.

Scotland has already experienced the impact of such operations. A study by disinformation firm Cyabra uncovered more than 1,300 Iranian-backed bot accounts on X (formerly Twitter) that were posting pro-independence, anti-Brexit, and pro-Tehran content using AI-generated images to create believable personas. These bots, which incorporated Scottish slang to appear authentic, potentially reached 224 million viewers before falling silent during the Iranian uprising.

“There’s a lot of evidence of efforts to subvert or manipulate public discourse,” said Dr. Paul Reilly, senior lecturer in communications, media and democracy at the University of Glasgow. These campaigns often aim to “sow confusion and discord” and “amplify polarization,” he explained.

Deepfakes—AI-generated audio, images, or videos that realistically depict people saying or doing things they never did—represent another significant threat. While initially crude, this technology has advanced dramatically in recent months, raising serious concerns about its potential to deceive voters.

Just 18 months ago, Dr. Reilly would have considered the threat of AI-generated media to the democratic process an exaggeration. However, the proliferation of realistic deepfakes targeting political figures has made it “a big concern” for upcoming elections.

Several Scottish politicians have already been targeted. First Minister John Swinney appeared in a viral fabricated clip claiming the SNP would “do whatever we want” after an election. Another fake video of Green MSP Maggie Chapman, watermarked with the Scottish Parliament TV logo, showed her falsely advocating for the abolition of Scotland’s roads.

AI chatbots present yet another challenge as voters increasingly turn to them for political information and advice. A survey by the AI Security Institute found that 32 percent of chatbot users—equivalent to 13 percent of all eligible UK voters—reported using these tools to inform their electoral choices.

“This is a huge figure, given how AI is still relatively new,” Reilly noted, adding that AI assistants are “shaping how elections are won and lost… which is quite alarming.”

Research from MIT found chatbots more persuasive than traditional campaign materials and even experienced campaigners. However, these systems have also been found to spread climate conspiracy theories, mimic Russian disinformation, and misrepresent news content.

At a January press briefing in Edinburgh, the Electoral Commission identified AI—particularly deepfakes—as one of its biggest concerns ahead of the Holyrood election, alongside candidate safety. The commission is working with the Home Office on a pilot program using software to detect deepfakes and will tackle fake media targeting candidates when it falls within its remit.

“We closely monitor how deepfakes are used in campaigning in the UK and overseas and are aware of the risks they pose to election security and voter confidence,” a commission spokesperson said.

While the UK’s Online Safety Act includes provisions against foreign interference that could address bot swarms in certain circumstances, regulatory gaps remain. The law prohibits making false statements about a candidate’s personal character to influence election results, but false statements about other matters aren’t necessarily illegal.

Dr. Reilly argues that all institutions share responsibility for combating misleading content, including media outlets that can fact-check claims and social media companies, which he believes “are still not taking their roles seriously enough.”

As Election Day approaches, officials are urging voters to think critically about campaign material and verify information before sharing it. In an era of declining trust in institutions, the public’s ability to identify AI-generated content and question the authenticity of online sources has become essential to preserving electoral integrity.

“It’s an information crisis,” Reilly concluded, “and the lack of trust in institutions is something which has been brewing for well over a decade.”