Deepfakes and disinformation in Pakistan: Politicians’ Perceptions and the Politics of Regulation

The emergence of deepfake technology, synthetic audiovisual media generated through advanced deep learning, has sparked debates among scholars regarding its potential for political, ethical, and social harm. Concerns have been raised about an impending “information apocalypse,” the erosion of trust in media, and the undermining of confidence in political institutions. As deepfakes become increasingly sophisticated and widespread, they pose new challenges for governance, political communication, and regulatory control. This study explores how deepfakes shape Pakistan’s evolving disinformation landscape and examines their role in shaping political narratives, influencing institutional credibility, and driving regulatory debates. This study explores these dynamics through two key research questions: (a) How are deepfakes perceived by Pakistani politicians within the broader socio-political landscape? (b) What are the implications of these perceptions for power dynamics, governance, and regulatory responses? To address these questions, semi-structured interviews were conducted with 25 Pakistani politicians. Participants were selected from national and provincial levels. Thematic analysis of anonymized transcripts identified recurring patterns in perceptions of deepfakes, institutional distrust, and regulatory preferences. Findings show that Pakistani politicians largely frame deepfakes as both a continuation and amplification of traditional disinformation tactics. While the current use of deepfake technology remains limited, respondents widely anticipate its potential to further fragment the information ecosystem. Politicians highlight the ease with which audiovisual content can be manipulated and rapidly disseminated, posing a significant threat to the veracity of public discourse. In Pakistan’s media landscape marked by partisan reporting, declining institutional trust, and the erosion of traditional gatekeeping mechanisms, deepfakes are expected to exploit and intensify existing vulnerabilities. Respondents describe deepfakes as persuasive not only because they “look real,” but because they travel through virality, polarization, and pre-existing reputational assumptions, often mixing truth with fabrication and leveraging morally charged domains such as religion and gender. Many respondents also predict that the weaponization of deepfakes will allow political elites to pre-emptively dismiss incriminating content by attributing it to fabrication, thereby undermining accountability and exacerbating political polarization. This study identifies that Pakistani politicians adopt a multifaceted approach to countering deepfakes, including verification practices and legal reforms, alongside growing demands for stronger accountability for those who circulate deepfakes. However, they also express concern that tighter regulation may expand state surveillance and censorship, potentially centralising political power. Overall, deepfakes are not understood as the root cause of political instability, but as a catalyst that magnifies pre-existing systemic problems. The study argues that anticipated deepfake proliferation is likely to deepen vulnerabilities in Pakistan’s information ecosystem and may reinforce authoritarian tendencies under the guise of digital regulation. This study contributes to academic discussions on deepfakes by providing a contextualized analysis of how they are perceived in a developing country. Furthermore, it adds to the broader discourse on the weaponization of AI enabled visual disinformation in environments already marked by media fragmentation and political polarization.