Rahul Dogra
As India prepares for the 2024 elections, there are growing fears about the impact of generative artificial intelligence technologies like deepfakes on the integrity of the democratic process. Experts have long warned that advances in AI would enable malicious actors to create hyper-realistic fake images, videos and audio that could mislead voters and sway election outcomes at scale.
Until recently, deepfakes were too expensive and crude to produce effectively. But this has changed drastically in recent years. Powerful generative AI tools can now churn out manipulated media, like hyper-realistic images/videos, within seconds at minimal cost.
Additionally, sophisticated generative AI tools can now create cloned human voices and hyper-realistic audio in seconds at minimal cost.
What if a celebrity like Sachin Tendulkar, Virat Kohli, or Amitabh Bachchan personally calls you and tells you to vote for a certain candidate?” A lot of people will listen to it as they idolize certain public figures but it is not actually a real voice but a cloned voice. When paired with social media algorithms, this fake AI-generated content – including cloned celebrity voices – can quickly target narrow audiences and potentially distort elections through widespread misinformation. The ability to create cloned voices that mimic public figures with perfect authenticity adds an insidious dimension to the AI threat for Indian democracy.
Using familiar, trusted names adds legitimacy and credibility, making the threat of cloned celebrity voices especially concerning. AI-powered data analytics and social media targeting will allow political parties to micro-target voters with extremely customized messaging and ads. This level of precise targeting could sway voters who are on the fence.
Deepfake threats loom large over India’s upcoming elections. Politicians and parties that harness generative AI stand to gain an advantage as manipulated content targeting opponents spreads rapidly online. However, the Election Commission currently lacks regulations around deepfakes or AI-manipulated campaign content. There are also no existing laws prohibiting the use of deepfakes in Indian elections, though the government is examining new policies for AI technologies.
Let us consider another scenario where a deepfake video surfaces of spiritual leader Sri Sri Ravi Shankar personally endorsing a particular candidate. Many of his devotees would likely be influenced to vote for that candidate, though the video was not actually of him but a cloned deepfake. The influence such spiritual figures have, especially when combined with widespread sharing of the manipulated video on social media platforms, makes the impact of such deepfakes difficult to counteract .
While most political consultancies likely advise against officially using deepfakes due to ethics, third parties and online trolls are expected to create and spread AI-manipulated content targeting candidates. Identifying authentic versus AI-generated content will prove difficult for voters.
More threats involve AI tools that analyze vast amounts of social media and online data will give parties deep insights into public sentiment, voter preferences and potential turnout. They can then tailor their campaigns accordingly to maximize support. AI bots could be deployed at scale to spread favorable content, attack opponents and astroturf public opinion. The anonymity of bots makes it hard to trace and regulate their influence.
Potential solutions to mitigate deepfake threats include banning their creation/spreading during elections, requiring social media platforms to label manipulated political content to educate voters, implementing policies within parties avoiding AI-generated content, launching voter awareness campaigns, incorporating cybersecurity/fact-checking into school curriculums, mandating transparency around political ads to identify entities behind them, and developing digital literacy initiatives to improve skills in verifying information and detecting online manipulation.
How stakeholders in India address the challenge of deepfakes for democracy will shape the future of Indian elections. With generative deepfakes now posing a serious risk, India must act swiftly and decisively to regulate their use while empowering citizens with knowledge and tools to identify manipulated media.
The time to act is now in order to maximize AI’s benefits for democracy while minimizing its potential for harm. A holistic strategy is needed involving policies, regulation, industry practices, education and awareness initiatives. Without concerted action, deepfakes threaten to undermine election integrity, spread mass disinformation and erode public trust in India’s democratic institutions. Democracy itself may be at risk if India fails to effectively confront this growing challenge for democratic governance in the AI era.
Follow our WhatsApp channel
