Turing research discounts deepfake influence on elections as govts ramp up regulation

The year 2024 will be remembered for several globally important national elections but also for warnings over the deepfake misinformation threat. Researchers, however, are uncovering that the promised deepfake wave has turned out differently than expected – but this is not stopping governments from bringing new regulations to reign it in.

New research by the UK-based Alan Turing Institute showed that there is no evidence that AI-generated misinformation impacted recent elections in the UK, France or the European Union. The institute identified just 16 viral cases of AI disinformation or deepfakes during the UK general election while only 11 viral cases were recorded in France and the EU.

Concerns remain about the misinterpretation of parody and satire deepfakes as well as the rising number of deepfake pornography images targeting women.

“Echoing our previous report, there remains no evidence AI has impacted the result of an election, but we remain concerned about the persistent erosion of confidence in what is real and what is fake across our online spaces,” says Sam Stockwell, lead author and research associate at the Turing Institute.

While the Institute announced more research results in November, the conclusions echo experience from India, which held the world’s largest elections between April and June. While millions of Indian voters viewed deepfakes on their social feeds, much of it was created by political parties to connect to voters instead of misinforming them, according to researchers from Harvard Kennedy School.

The Turing Institute notes that AI can even be beneficial to the democratic process when it’s used for clever marketing campaigns. Similar conclusions may have been reached by the United Nations Secretary-General António Guterres in its message for the International Day of Democracy last Sunday.

“Left unchecked, the dangers posed by artificial intelligence could have serious implications for democracy, peace, and stability. This can start with the proliferation of mis- and disinformation, the spread of hate speech and the use of so-called deepfakes,” says Guterres  “Yet, AI has the potential to promote and enhance full and active public participation, equality, security, and human development.”

Countries using AI for good governance could create more transparent and accountable systems, he added, highlighting the conclusion from the recent report of the UN’s High-Level Advisory Body on AI.

Countries continue to regulate deepfakes

Despite good news for the 2024 global election cycle, authorities are warning that governments are still not safe from the deepfake threat.

Last week, Interpol warned that the Philippines should develop regulations on using deepfakes to minimize political deception during the 2025 general elections. The international organization for police cooperation argued that deepfakes could be used to swing elections toward certain candidates, news outlet Philstar reports.

The Pacific country has already been a victim of a mass disinformation campaign during national elections in 2022, while this year, deepfake content was used to target President Ferdinand Marcos Jr.

The U.S. is also making steps towards more stringent laws against deepfakes, albeit on a state-by-state level. Earlier this year, Colorado brought legislation requiring campaign ads to mark AI-generated content. California is now making similar moves.

This week, Californian Governor Gavin Newsom signed three measures against deepfakes and other AI content. This includes removing deceptive and digitally altered or created content from large online platforms during elections and disclosing AI-generated content in electoral advertisements. Another legislation will protect the digital likeness of actors and performers by ensuring their voice or likeness cannot be without consent.

Last week, an AI-generated image of pop star Taylor Swift seemingly endorsing former President Donald Trump appeared on social media, prompting fresh warnings about the dangers of deepfakes from cybersecurity experts, NBC reports.

Singapore, on the other hand, is considering legislation that allows candidates to flag deepfakes of themselves, according to the South China Morning Post.

Other jurisdictions are also introducing legislation against AI-generated content including the government of South Australia which plans to introduce a bill against the creation and distribution of non-consensual sexually explicit deepfakes. According to the Australian eSafety Commissioner, explicit deepfakes increased as much as 550 percent year on year since 2019.

China was the first country to introduce legislation against AI-generated content in January 2022.

Article Topics

California  |  deepfakes  |  elections  |  fraud prevention  |  generative AI  |  legislation  |  Turing Institute  |  United Nations