AI-generated 'victory speech' by jailed Imran Khan sparks ethical debate in Pakistan's elections (WATCH)
According to The New York Times, Imran Khan's Pakistan Tehreek-e-Insaf (PTI) party utilized AI technology to simulate the former Pakistan PM's voice based on notes he transmitted from jail.
Former Prime Minister of Pakistan, Imran Khan, currently jailed on what his supporters claim are politically motivated charges, seemingly delivered a triumphant address Saturday night following his party's success in the country's general election. However, the speech wasn't delivered by Khan himself but rather by an artificial intelligence (AI) mimic of his voice.
According to The New York Times, Khan's Pakistan Tehreek-e-Insaf (PTI) party utilized AI technology to simulate Khan's voice based on notes he transmitted from jail. The video, showcasing Khan and his supporters, featured audio of his voice claiming victory and alleging electoral fraud by opponents, alongside a disclaimer crafted with AI.
This development highlights the potential of AI to circumvent speech suppression and the technology's ability to deceive. Khan, despite being incarcerated, has been rallying supporters throughout the election using online platforms and now AI to convey his message, despite censorship efforts by Pakistan's military leadership.
AI expert Toby Walsh, as cited by The New York Times, remarked on the irony of the situation, acknowledging the potential positive outcome of enabling someone falsely accused of corruption to communicate with their supporters. However, he also emphasized the erosion of trust in visual and auditory information due to such developments.
"In this case, it's for a good end, perhaps an end we'd support - someone who's locked up on trumped-up charges of corruption being able to speak to his supporters," said Walsh. "But at the same time, it's undermining our belief in the things we see and hear."
The utilization of AI to replicate Khan's voice adds to a growing trend of AI integration in political campaigns. For instance, South Korean President Yoon Suk Yeol's campaign employed an AI avatar for promotional purposes. Additionally, recent incidents, like the AI-generated voice mimicking US President Biden in a voter suppression robocall in New Hampshire, prompted the FCC to ban AI-generated voices in robocalls.
Khan's AI-driven victory speech underscores the evolution of political communication strategies in the digital age. Despite being barred from running himself, Khan's party adopted a tech-centric approach, utilizing online rallies and social media to reach supporters amidst censorship challenges.
While AI offers unprecedented opportunities for communication and engagement, its misuse raises significant ethical and societal concerns. As political campaigns increasingly rely on AI-generated content, regulators and policymakers face the daunting task of balancing innovation with the preservation of democratic principles and the integrity of public discourse.