Fastag may go tagless with AI, says DPI architect

BENGALURU: India’s next-generation Fastag system may not require a physical tag at all. Vehicles could be identified through roadside cameras, their type and number detected in real time, and tolls deducted automatically without a windshield sticker, Pramod Varma, a key architect of India’s digital public infrastructure (DPI), said on Tuesday.Speaking at the ConvoZen Conversational AI Summit in Bengaluru, Varma said such computer vision-led systems represent the next layer of India’s digital evolution, where artificial intelligence is embedded directly into public infrastructure and operates at the edge. The proposed system, he said, would process thousands of vehicles passing through highways each minute, detecting and charging them instantly.The broader shift, he argued, is from building digital identity to enabling “digital agency.” Over the past decade, India formalised nearly a billion people into the economy through platforms such as Aadhaar, UPI, DigiLocker and account aggregators. Citing global studies, Varma said India achieved in seven to eight years what might otherwise have taken five decades.“The last decade formalised a billion people. The next decade must meet their aspirations,” he said, outlining a move toward livelihood, skilling and income-generating transactions powered by AI.A key barrier now is interface complexity. While about 550-600 million Indians actively use digital systems, a large share of the adult population still struggles with forms and text-heavy interfaces. Conversational AI, particularly voice systems that allow broken, code-mixed language, could lower that barrier.Varma pointed to a demonstration under the India Energy Stack, where a farmer seeking to increase income could state that goal, after which an AI system analysed pricing, structured an energy trade contract and executed payments through UPI. “The real breakthrough is complexity reduction,” he said.He also cautioned that AI-driven fraud and deepfake attacks make “proof of personhood” and cryptographically verifiable digital trust essential in the AI era, adding that identity frameworks will need to extend to digital agents and machines.