Regulating AI Deepfakes in 2024 is the Biggest Challenge for Countries

Deepfakes, which are highly realistic fabricated videos or images, can deceive voters, manipulate public opinion, and tarnish the reputation of political parties and politicians

More than four billion people worldwide will take part in elections this year, including in the US, UK, India, and South Africa. However, the increasing prevalence of disinformation, fuelled by the growing use of deepfakes and generative AI, means that politicians and voters alike will need to be on high alert.

In the 2024 Global Risks Report (https://apo-opa.co/3wRJRWs), the World Economic Forum identified AI-generated misinformation and disinformation as the second most significant global risk after extreme weather. This concern is well-founded, especially with the rapid rise of tools like generative AI-empowered deepfakes, posing a threat to the integrity of democratic processes. Deepfakes, which are highly realistic fabricated videos or images, can deceive voters, manipulate public opinion, and tarnish the reputation of political parties and politicians.

Anna Collard, SVP Content Strategy and Evangelist at KnowBe4 AFRICA, a cybersecurity training organisation, says there has been a vast improvement in the quality of deepfakes over the past year, making them more convincing and difficult to identify. In a notable incident from October last year, cybercriminals used deepfake videos (https://apo-opa.co/4aewZYY) to impersonate African Union Chairperson Moussa Faki Mahamat during online conversations with European diplomats. Similarly, in December, Facebook removed over 100 deepfake paid advertisements (https://apo-opa.co/3TBfnkt) featuring British Prime Minister Rishi Sunak after nearly half a million people viewed them.

“Deepfakes pose major risks in the run-up to elections, both in the UK and South Africa,” says Collard. “Voters need to be aware of this risk and be proactive about mitigating their impact.”

Top News on AI Deepfakes:

McAfee Launches Deepfakes Lab to Thwart Video Election Disinformation

Be Aware of Deepfakes

According to Collard, the initial step in combating the impact of misinformation and disinformation through deepfakes is raising awareness. “It’s crucial for the public to be aware of the existence of this phenomenon,” she emphasises, adding, “The production of deepfake videos and images is cheap and easy.”

CounterCloud, a research project to showcase how easy it is to create fully automated (https://apo-opa.co/3wSNhZ0) disinformation campaigns, operated with a modest budget of only $400. Its goal was to counter negative depictions of the US propagated by Russian media. “If you’re a politician, expect deepfakes impersonating you to appear in order to sway public opinion,” she asserts. “Similarly, voters should not believe everything they see or hear on social media to be true.”

Verify information

Apart from raising awareness, another crucial principle is to verify information. “If a politician purportedly said something that’s deeply polarising, it’s important to verify whether they genuinely said it.” To counter the risk of deepfakes, the Independent Electoral Commission (IEC), in collaboration with Media Monitoring Africa, has launched an initiative called Padre (https://apo-opa.co/3THIwux). This initiative allows voters to fact-check information regarding South African political parties and their recent statements.

Google’s reverse image search (https://images.Google.com/) is another helpful tool that enables you to check if an online image has been used elsewhere. Additionally, FotoForensics (https://FotoForensics.com/) provides a more in-depth analysis of images. However, despite these tools, verifying information remains challenging. “Deepfakes are becoming more convincing, making verifying information even harder,” states Collard.

Don’t over-react

Lastly, when faced with inflammatory content, Collard stresses the importance of voters not overreacting. “Scammers want to whip up your emotions and get you in a state where you’re unable to think rationally,” she says. “Stay calm and don’t share content that is overly emotive. Whenever your emotions are triggered, slow down and verify, -it could be a hoax.”

Collard believes that it is every voter’s responsibility to educate themselves about the dangers of deepfakes. “AI-generated fake content has huge implications for society, especially during election periods,” she says. “To combat this form of disinformation and safeguard the democratic process, social media platforms, political parties and independent watchdogs as well as us the public will all need to work together.”

[To share your insights with us as part of editorial or sponsored content, please write to sghosh@martechseries.com]