As the election day dawns, voters eagerly queue up with their election cards in hand, only to be met with a surreal sight: cancelled elections, rendering their votes obsolete. This eerie scene flickers across your phone screen like a haunting snapshot, serving as a quirky yet ominous prelude to the deeper issues at hand — the rise of Artificial Intelligence (“AI”) -generated false news and the looming threat of deepfakes in elections. In light of these challenges, we must examine the role of AI in elections: is it a threat or can serve as an ally? This article delves into the complexities of AI’s impact on the electoral process, exploring both its potential to revolutionize and its capacity to disrupt.
AI Reaches Delhi High Court
The AI has not only entered the realm of election but has also reached the High Court of Delhi. In Lawyers’s Voice v Union of India & Ors, the petitioner sought a direction from the Election Commission of India (“ECI”) to formulate guidelines for verification of video messages to peremptorily weed out deep fake videos or for prohibiting the use of deep fake technologies in any media or mode of mass communication that may be employed for reaching out to voters, under the enforcement of the Model Code of Conduct and until the declaration of election results. Additionally, the PIL called for the ECI to gather data on deepfake technology usage related to political figures and submit a report to the court.
Furthermore, it demanded that Google, Meta, and X remove and block deepfake content or political figures from their social media platforms until election results are declared. These directions were sought based on instances of deepfake videos such as an AI-generated video indicating Congress leader Rahul Gandhi to be swearing-in as the Prime Minister of India or a deepfake video of Bollywood actor Aamir Khan seeking votes for Congress.
However, the High Court dismissed the PIL while observing that it cannot pass a direction in the middle of the elections and that the ECI is not remediless and will act on the issue. The court remarked that there is no technology under which uploading of deepfake videos can be stopped and “it can be created by anyone. Once you know the handle, the account is disabled. Once they disable the account, that’s max that can be done,”.
This judgment raises deeper question about the use of AI in election and its uncontrollable power of spreading fake news. In fact, AI is a big concern in elections not just in India, but also in the US and South Korea. Recently, an AI-generated voice claiming to be President Joe Biden circulated, urging voters not to participate and falsely suggesting they should save their vote for the general election in November. This misled voters into thinking Biden himself recorded the message.
AI- A Threat To Democracy?
The incorporation of AI into electoral processes gives rise to ethical dilemmas concerning privacy, transparency, and equity. AI algorithms, while aiming for impartiality, could inadvertently perpetuate biases, potentially resulting in discriminatory treatment towards specific voter demographics.
Furthermore, the opacity surrounding AI decision-making processes poses a threat to public confidence in the integrity of election results. Additionally, unequal access to resources among political parties may lead to an imbalance, jeopardizing the fairness of elections. Thus, the intersection of AI in elections acts as a threat to democracy and calls for immediate action, an attempt of which was made by the Ministry of Electronics and Information Technology.
MeitY’s Guidelines On Use Of AI In Elections
Recently, the Indian Ministry of Information Technology issued a significant advisory ahead of the Lok Sabha elections, targeting generative AI platforms like Google and OpenAI. These guidelines mandated that AI systems operated by companies such as Google and OpenAI must comply with Indian laws and refrain from actions that could compromise the integrity of the electoral process. Central to the guidelines is the requirement for platforms offering AI services in India to obtain explicit permission before deploying under-tested or unreliable AI models. Moreover, they must transparently disclose the potential fallibility or unreliability of AI-generated outputs to users. This measure aims to manage expectations and mitigate risks associated with misinformation or biased content inadvertently generated by these systems. Another pivotal aspect of the guidelines is the implementation of traceability measures. Platforms are instructed to embed unique metadata or identifiers in AI-generated content, enabling authorities to trace back and identify the originator of any misinformation or deepfake. This step is crucial in combating the spread of misleading information during sensitive periods like electoral campaigns.
Minister Rajeev Chandrasekhar emphasized that while the advisory is not legally binding, it sets a precedent for forthcoming legislative actions designed to impose stricter oversight on AI technologies. He underscored the global concern over AI’s potential misuse in influencing electoral outcomes, citing international instances where AI-generated content has been leveraged for malicious purposes.
These guidelines reflect a proactive approach by the Indian government to balance technological innovation with regulatory safeguards, particularly in sectors as impactful as AI. As India prepares for elections, ensuring AI systems uphold ethical standards and align with national interests becomes paramount. The move signals a maturing regulatory landscape where policymakers are increasingly vigilant about the societal implications of AI advancements, aiming to foster a responsible and accountable AI ecosystem in the country.
Conclusion
In conclusion, the integration of AI into electoral processes presents both promise and peril. While AI holds potential to enhance efficiency and engagement, its unchecked use poses significant threats such as the proliferation of deepfakes and misinformation. The recent guidelines from India’s Ministry of Information Technology are a crucial step towards mitigating these risks by enforcing transparency, permission-based deployment, and traceability of AI-generated content. Looking ahead, a comprehensive legislative framework is imperative. Such legislation should not only govern AI’s role in elections but also address broader issues of fake news dissemination. This would entail robust mechanisms for verifying AI-generated content, stringent penalties for misuse, and equitable access to AI technologies across political spectrums.
Moreover, international collaboration is essential to develop norms and standards that transcend borders, given AI’s global impact on democratic processes. By fostering responsible AI practices and ensuring adherence to ethical standards, policymakers can uphold electoral integrity and public trust in democracies worldwide. Ultimately, harnessing AI’s potential as a digital ally while mitigating its risks demands a concerted effort to strike the right balance between innovation and safeguarding democratic principles.
Authors: Muskkaan Verma & Bhumika Sharma
Contributor: Paavanta Arya