IIT Delhi and Amazon launch Raksham AI hackathon on deepfakes

Raksham is a small program in dollar terms, but it is symbolically important: one of India’s top engineering schools, the national cybercrime agency and a global tech giant are explicitly framing AI as both the problem and the solution for deepfakes and online fraud. By focusing student attention on voice cloning, fake media and financial scams, the hackathon is likely to generate a crop of niche tools and models that will quickly find their way into law enforcement and fintech workflows.

From an AGI perspective, the interesting piece is not the competition itself but the ecosystem it signals. India is effectively training a generation of engineers to think of generative models as part of the country’s cyber defense infrastructure. That aligns with similar moves in the U.S. and Europe and suggests that high end model access will increasingly be justified not only on productivity grounds but also on national security and consumer protection.

The more that governments treat AI as critical infrastructure for both offense and defense in the information space, the harder it will be politically to slow core capability work, even if safety concerns grow. Labs that can credibly position their models as tools against fraud, misinformation and abuse will have an easier time arguing for privileged treatment in policy and procurement.