South Asia is currently undergoing an important year, with Bangladesh and Pakistan having completed their elections, while Sri Lanka, Maldives, and India are preparing for theirs. Artificial intelligence (AI) is revolutionising political processes in a number of ways, such as increasing transparency, facilitating decision-making, and detecting bias. At the same time, AI also presents risks such as manipulation through deepfakes and the spread of misinformation. The use of AI in the elections of Bangladesh, Pakistan, and India has raised concerns about prospective violations of human rights. These include the rights to information; privacy; free and fair elections; and political engagement. This article explores the intersection of AI and human rights in the context of elections in South Asia and emphasises the significance of balancing technological advancements with the protection of democratic principles.

Elections in South Asia

Pakistan

In the recent parliamentary elections in Pakistan, AI-powered technology had both positive and negative impacts on the electoral process. The political parties of the country, particularly Pakistan Tehreek-e-Insaf (PTI), effectively employed AI to address the many obstacles they faced prior to the election. For instance, despite severe restrictions on their ability to organise, the party managed to productively host virtual gatherings and connect with the voters through AI-generated speeches of the imprisoned leader, Imran Khan, assuring his continued public presence. This image of him as a persecuted hero, facilitated by AI, was intended to boost solidarity and foster optimism among his voters. Moreover, the party introduced a digital portal to help voters identify the party’s candidates and electoral symbols in their constituencies in the absence of a single party electoral symbol.

However, the elections also revealed the negative aspects of deepfake content. Some users exploited AI-generated videos and audio messages to spread misinformation and manipulate voter behaviour, favouring and opposing parties and candidates. False announcements of election boycotts and withdrawal of candidates circulated on social media platforms causing uncertainty and disrupting the election.  This spread of misinformation exacerbated political tensions and undermined the credibility of the election process.  The implications of such actions are not just limited to elections but also threaten the core democratic values of transparency, accountability, and informed decision-making.

Bangladesh

Similarly, AI significantly influenced the election earlier this year in Bangladesh. Political parties utilised AI creatively to improve their digital campaign strategies. For instance, the ruling Awami League party strengthened its political dominance and controlled the narrative through the persecution of opposition parties, media outlets, and civil society. Such tactics undermine democratic values of free speech, access to information and free and fair elections. deepfake technology was used to spread misinformation and disrepute opposition leaders. Controlling online discourse helped the party maintain power and manipulate public opinion.

Such tactics weaken political diversity and the public’s ability to make fully informed choices in elections. The long-term consequences may include reduced public faith in democratic institutions, dissatisfaction with the political system, and low voter turnout in future elections. In addition, the online presence of political parties influenced by their financial resources contributes to campaign disparity. Indeed, parties with greater resources can leverage more advanced algorithms to target certain populations and spin narratives in ways that benefit them at the expense of a free and fair election.

India

In the ongoing election campaign in India, the two major political parties, the Bharatiya Janata Party (BJP) and the Indian National Congress (INC), are making strategic use of AI to influence voter perceptions and achieve electoral benefits. They actively recruit social media influencers capitalising on AI to create attention-grabbing content, such as first-hand accounts to engage voters and maintain a positive public image. For instance, Sharvan Patel, a Rajasthan-based influencer with over 470,000 followers, allegedly partnered with the INC to promote their policies. His Instagram post, which received more than 4 million views, showed a woman receiving a smartphone from the state’s Chief Minister, implying direct material benefits from the party. In another instance, Deepti Maheshwari of BJP used a network of 1000 nano-influencers to deal with negative press about their candidate selections. Such examples raise concerns regarding the spread of misinformation, hate speech, and paid endorsements, highlighting the need for regulations to control the use of AI in political campaigning.

The unregulated employment of such approaches undermines voters’ right to information and prevents them from making rational choices. deepfakes may harm reputations and violate privacy. This is further exacerbated by South Asia’s digital infrastructures and the urban-rural divide with inadequate connectivity and access issues. This divide contributes to inequality, especially among the marginalised populations which may lack the access to technology, limiting their political engagement.

South Asia’s digital infrastructures

South Asia faces substantial disparities in digital infrastructure. Urban areas have better internet connectivity due to concentrated investments and denser populations. However, rural areas are neglected due to a lack of infrastructure development and connectivity issues. This digital divide creates a serious information gap. Without access to the necessary digital skills and resources, these communities become easy targets for misinformation. The issue is further amplified by realistic fake content created by AI, making it harder to distinguish truth from fiction. Furthermore, this can result in creation of ‘echo chambers,’ where people are only exposed to the information that further reinforces their existing biased views, limiting their ability to assess the information and situations critically.

The governments in South Asia have the capacity to use AI-powered technology to exert control over the flow of information, particularly during elections. This solidifies concerns that AI algorithms could be used to censor content, silence dissent, and manipulate public opinion, raising questions regarding the legitimacy of elections. Examples exist of censorship during elections in Pakistan and Bangladesh, where the governments restricted social media and news channels. The potential of AI to automate such practices is alarming. Further complicating matters is the uneven quality and reliability of digital infrastructure across South Asia. While some areas boast high-speed internet, others struggle with frequent outages and limited bandwidth. This inconsistency cripples the effectiveness of AI tools in monitoring elections and voter outreach.

Addressing the divide and increasing connectivity

To address the digital divide and reduce susceptibility to AI violation, several steps may be implemented to build a more inclusive and durable digital infrastructure empowering individuals and protecting their rights. These include enhancing Internet connectivity, bridging the infrastructure gaps by promoting infrastructure sharing, increasing affordability, and utilising Universal Services and Access Funds (USAFs) effectively. USAFs serve as a bridge to ensure that funds are directed towards the programs that improve services for citizens while also creating new opportunities for telecommunications and content providers. The ultimate goal is for the services to be profitable and self-sustaining.

Achieving this goal requires collaborative efforts, involving both public and private sectors and civil societies. It includes development of effective policies, partnerships, ensuring transparency and accountability, and promoting cooperation. In addition, digital education is also important through tutorials and guidance to improve digital literacy. Furthermore, ensuring that people have the ability to understand and analyse the source of content is important. This can be done by the inclusion of Web Content Accessibility Guidelines (WCAGs) in domestic policy. Finally, upgrading legal and regulatory frameworks to reflect technological advancements and new opportunities in the digital economy is essential for adopting digital inclusion and protecting individuals’ rights.

Regulatory framework for AI governance

Currently, South Asian countries confront issues related to digital infrastructure literacy and socioeconomic gaps which increase the potential for AI’s misuse. Hence, effective regulations are necessary to prevent such risks.

Earlier this month, the European Union (EU) enacted the world’s first major legislation for AI in media and technology. The Artificial Intelligence Act prohibits risky AI uses; mandates transparency and copyright compliance for high-risk AI; and requires labelling of manipulated media. South Asian countries can also make efforts to develop similar legislation, tailoring it to their own specific contexts based on their legal and political systems.

To create and implement such legislation, internationally recognised standards may be used as a foundation for regulation, model laws for lawmakers, and regulatory standards when included by reference. Standards built on expertise from public and private industries and civic societies are more flexible and adaptable compared to rigid regulations and allow continuous innovation and development in technology.

South Asian countries can empower their citizens by adopting legislative acts similar to that of the EU. This legislation would establish an umbrella for basic rights and the ethical use of AI. It will include protections for AI systems, the prohibition of applications that threaten freedom and rights, explicitly drafted guidelines for high-risk AI, and mechanisms to guarantee transparency and accountability. It will lead to a future where AI will only be a tool for progress. Implementing regulations on the use of AI in electoral processes could prove beneficial for not just democracy but also in boosting the global reputation, while also increasing innovation and addressing the digital gap.

Conclusion

As AI becomes prevalent in electoral processes around the world, it is becoming equally important to maintain a balance between innovation and the rights of voters. Elections in South Asia highlight both the benefits and risks associated with the use of AI and emphasise the need for effective regulations. By adopting such legislation, the countries can safeguard electoral integrity, foster innovation and promote cooperation. Collaborative efforts are key to developing and implementing these regulations effectively and protecting democratic principles and human rights in this digital era.

About the authors

Aiman Iqbal is the managing editor of the Journal of Peace and Diplomacy, published by the Institute of Peace and Diplomatic Studies (IPDS), and the assistant editor of The Diplomatic Insight magazine. She holds an MPhil in international relations. She can be reached at aimaniqbal106@gmail.com

Shireen Mushtaq is an Assistant Professor in the International Relations Department at Kinnaird College for Women, Lahore. She holds a Ph.D. in international relations and specialises in South Asian regional dynamics, including nationalism, security, and foreign policy. She can be reached at shireen.mushtaq@kinnaird.edu.pk

Centre for Human Rights (CHR) blog