NEW DELHI (AP) — Bollywood stars rarely influence politics, so videos showing two celebrities criticizing Indians Prime Minister Narendra Modi – and endorsing their main opposition, the Congress Party – were bound to go viral.
But the clips A-list actors Aamir Khan and Ranveer Singh were fake AI-generated videos that were yet another example of false or misleading claims circulating online with the aim of influencing Elections in India. Both parties filed complaints with the police, but such actions do little to stem the flow of such misinformation.
Claims circulating online in India recently distorted details about votingclaimed without evidence that the election will be rigged and asked violence against India’s Muslims.
Researchers who track misinformation and hate speech in India say technology companies’ misapplication of their own policies has created the perfect conditions for harmful content that can distort public opinion, encourage violence and leave millions of voters wondering what to believe.
“Does an undiscerning user or an ordinary user have no idea if it is someone, an individual sharing their thoughts on the other side or if it is a bot?” Rekha Singh, a 49-year-old voter, told the Associated Press. Singh said she worries that social media algorithms distort voters’ view of reality. “Then you are biased without even realizing it,” she said.
In a year full of big elections, the Voting expanding in India stands out. O the most populous country in the world has dozens of languages, the largest number of WhatsApp users and also the largest number of YouTube subscribers. Almost 1 billion voters can vote in the elections, which runs until June.
Tech companies like Google and Meta, which owns Facebook, WhatsApp and Instagram, say they are working to combat misleading or hateful content while helping voters find trustworthy sources. But researchers who have long tracked disinformation in India say their promises ring hollow after years of failed enforcement and “one-size-fits-all” approaches that fail to take into account linguistic, religious, geographic and cultural diversity.
Given India’s size and its importance to social media companies, one would expect greater focus, say disinformation researchers who focus on India.
“Platforms are making money from this. They are benefiting from this and the whole country is paying the price,” said Ritumbra Manuvie, a law professor at the University of Groningen in the Netherlands. Manuvie is the leader of The London Story, an Indian diaspora group that last month organized a protest outside Meta’s London offices.
Research by the group and another organization, India Civil Watch International, found that Meta allowed ads and political posts that contained anti-Muslim hate speech, Hindu nationalist narrativesmisogynistic posts about female candidates, as well as advertisements that encourage violence against political opponents.
The ads were viewed more than 65 million times in 90 days earlier this year. Together, they cost more than $1 million.
Goal defends your work in global elections and disputed the investigation’s findings on India, noting that it expanded its work with independent fact-checking organizations before the election, and has employees around the world ready to take action if its platforms are misused to spread disinformation. Nick Clegg, president of global affairs at Meta, said of India’s election: “It’s a huge, huge test for us.”
“We have months and months and months of preparation in India,” he told The Associated Press during a recent interview. “We have teams working 24 hours a day. We have multilingual fact checkers operating in India. We have a 24-hour escalation system.”
YouTube is another problematic misinformation site in India, experts say. To test how well the video-sharing platform enforced its own rules, researchers from nonprofits Global Witness and Access Now created 48 fake ads in English, Hindi and Telugu with false information about voting or calls for violence. One claimed that India had raised the voting age to 21, although it remains 18, while another claimed that women could vote by text message, although they cannot. A third called for the use of force at polling places.
When Global Witness submitted the ads to YouTube for approval, the response was disappointing, said Henry Peck, a researcher at Global Witness.
“YouTube didn’t act on any of them,” Peck said, and instead approved the ads.
Google, which owns YouTube, criticized the research and noted that it has several procedures in place to detect ads that violate its rules. Global Witness removed the ads before they could be detected and blocked, the company said.
“Our policies explicitly prohibit ads that make demonstrably false claims that could undermine participation or confidence in an election, which we enforce in several Indian languages,” Google said in a statement. The company also highlighted its partnerships with fact-checking groups.
AI is this year’s newest threat, as program advancements make it easier than ever to create realistic images, videos, or audio. AI deepfakes are emerging in elections around the world, from Moldova to Bangladesh.
Senthil Nayagam, founder of an AI startup called Muonium AI, believes there is a growing demand for deepfakes, especially from politicians. In the run-up to the election, he received several questions about how to make political videos using AI. “There’s a market for it, without a doubt,” he said.
Some of the forgeries that Nayagam produces feature dead politicians and should not be taken seriously, but other forgeries circulating online can potentially mislead voters. It’s a danger Modi he highlighted it himself.
“We need to educate people about artificial intelligence and deepfakes, how it works, what it can do,” Modi said.
India’s Ministry of Information and Technology has instructed social media companies to remove misinformation, especially deepfakes. But experts say the lack of clear regulation or law focused on AI and deepfakes makes it harder to eliminate them, leaving voters to determine what is truth and what is fiction.
For Ankita Jasra, 18, who voted for the first time, these uncertainties can make it difficult to know what to believe.
“If I don’t know if what’s being said is true, I don’t think I can trust the people who run my country,” she said.
___
AP journalists Matt O’Brien in Providence, Rhode Island, and Rishi Lekhi in New Delhi contributed to this report.
