Key facts
- Small, open-source AI models identified as key deepfake enablers in India
- These models run on consumer hardware with minimal technical skill
- Deepfakes in India used for political disinformation, scams, and harassment
- Regulatory focus on big AI labs may miss the actual threat vector
Business Standard reports that smaller AI models, rather than cutting-edge systems from major labs, are likely the key technology enabling India's escalating deepfake crisis. These models are lightweight enough to run on ordinary consumer hardware, are freely downloadable from open-source repositories, and require minimal technical skill to deploy, making them far more accessible to a wide range of bad actors than previously assumed.
India has in recent years seen a sharp rise in deepfake incidents spanning political disinformation — including fabricated videos of senior leaders — non-consensual intimate imagery targeting women, and voice-cloning scams that have defrauded families by impersonating relatives. The low barrier to entry provided by small AI models means that producing convincing fakes no longer demands significant compute resources or specialised knowledge.
The finding complicates regulatory responses, since efforts focused on large AI companies or data centres would do little to address threats generated on personal devices using open-source tools. For everyday Indians, the implication is stark: deepfake-driven fraud and harassment are likely to become more frequent and harder to detect, raising urgent questions about digital literacy, platform accountability, and law enforcement capacity.