What do Bollywood deepfakes mean for global democracy?

Faked footage of Bollywood stars criticising Prime Minister Narendra Modi and urging people to vote for the opposition has been viewed more than half a million times as India goes to the polls.

The rapid spread of the AI-generated clips, in which Aamir Khan and Ranveer Singh "purportedly say Modi failed to keep campaign promises", underlines the potential role that deepfakes could play in India's "mammoth" six-week election, said Reuters – and in other votes around the world.

What did the commentators say?

Indians are highly engaged online. In a country of nearly 1 billion voters, the number of people with access to the internet is expected to surpass 900 million by 2025, according to Statista. The average Indian spends more than three hours a day on social media, said the Indian Institutes of Management.

Subscribe to The Week

Escape your echo chamber. Get the facts behind the news, plus analysis from multiple perspectives.

SUBSCRIBE & SAVE

Sign up for The Week's Free Newsletters

From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.

From our morning news briefing to a weekly Good News Newsletter, get the best of The Week delivered directly to your inbox.

Sign up

But this has opened the country up to mis- and disinformation. In the World Economic Forum's Global Risks Report for 2024, India topped the ranking of countries facing the highest risk from disinformation. And "for many seasoned India fact-checkers, this came as no surprise", said the Reuters Institute. 

In 2018, the BBC reported on research in India, Kenya and Nigeria into the way ordinary citizens engage with and spread fake news. "Facts were less important to some than the emotional desire to bolster national identity," the broadcaster said. Analysis of social media found that right-wing networks were "much more organised" than those on the left, "pushing nationalistic fake stories further". 

Part of the problem is the lack of clear guidelines in India around the use of AI, said Divyendra Singh Jadoun, a former student politician for a group affiliated with the country's opposition Congress Party, who now creates deepfake videos online. Speaking to Bloomberg, he said: "Deepfakes can easily sway the voter, as many are new to the internet, and get duped even by poor-quality ones."

India is "not unique in lacking regulation", said Bloomberg. "There are few guidelines and legal structures anywhere." And with more than 60 nations and the European Union due to go to the polls this year, "for the first time, potent breakthroughs in generative AI have the power to influence outcomes from Brussels to Delhi, Washington and beyond".

Large language models, which led to the creation of apps such as ChatGPT, have been the catalyst for "a flood of easy-to-use AI apps allowing single-person shops to churn out deepfakes at alarming speed".

The AI-generated deepfakes now being employed in elections across the world "are the sort of canaries in the coal mine of potentially what's to come in some of these elections", said Ginny Badanes, general manager Microsoft's Democracy Forward. "Whether that's going to have a direct impact on the election itself – that's obviously not something any of us can be really sure of."

What next?

Governments worldwide are grappling with the regulation of AI as they aim to balance the potential harms with the economic opportunities that such technology presents. 

Recent initiatives include non-binding agreements such as the Group of Seven principles on AI and the UK's AI Safety Summit conclusions. Others have taken "more concrete steps", said Bloomberg, with the EU planning mandatory rules and China implementing controls on AI use.

But sceptics have argued that these, mainly voluntary, commitments "don't go far enough at a time when companies are cutting their human trust and safety teams and outsourcing content moderation to AI".

"Many elections are decided on tiny percentile differences," Marietje Schaake, international policy director at the Stanford University Cyber Policy Center and a former member of the European Parliament, told Bloomberg. "You don't have to persuade significant amounts of people and can still have the bulk of the population saying, 'Oh, no, well, we were not at all impacted by it.'"