AI should not be given so freely to the general public

In the 2024 Indian general election, political parties utilized AI to deploy deepfakes and disinformation campaigns on an unprecedented scale. These technologies were used to generate realistic, albeit fabricated, content featuring prominent political figures to influence voter sentiment and shape the public agenda.

What happened

Political campaigns leveraged AI to create deceptive media that targeted opponents and bolstered their own candidates. In one instance, a deepfake video was circulated showing Prime Minister Narendra Modi dancing to a Bollywood song. Opposing parties also engaged in this practice; the Bharatiya Janata Party shared an AI-generated video depicting a jailed opponent playing guitar and singing, while the Congress Party created a video mocking Modi by altering song lyrics to imply he was handing the country over to large corporations.

Beyond video, AI was employed to create propaganda images and manipulate audio. Political parties used voice cloning technology to reach voters directly via personalized messages. Over 50 million AI-generated calls were placed in the two months leading up to the April elections, a method that was reportedly eight times cheaper than utilizing human-staffed call centers. Furthermore, AI helped campaigns bridge language barriers by generating content across India's 22 official languages and numerous regional dialects. The content was often designed to evoke strong emotional responses and was disseminated through less regulated platforms, frequently amplified by local content creators. This widespread use of AI-driven tactics raises significant concerns regarding the stability of democratic institutions and the potential for similar disruptions in future global elections.