Tech Explained: India’s tech infra as public good can be emulated for AI across Global South, says ITU official in Simple Terms

New Delhi: India deploying its massive technological infrastructure like Aadhaar and UPI as a public good instead of a for-profit commercial service serves as a highly replicable playbook for the Global South to drive AI diffusion, Fred Werner, chief of strategic engagement at UN International Telecommunications Union (ITU) told ET.

“If India can scale technology for over a billion people, it can be adapted for a country of two million,” Werner said, adding that the country strongly demonstrates the essential preconditions required to safely adopt and scale AI.

Werner leads the ITU’s AI For Good Summit that is working on creating standards that will be adapted by every player in the AI ecosystem.
He said India proves that a country does not necessarily need to possess the world’s largest language models or massive, centralised AI factories to be AI-ready. Its strength lies in AI diffusion and adoption.

He said the country is combining the speed of startups, scale of government, and the technical rigour of academia, working with ITU professionals in a secure sandbox environment where it uses open datasets from the UN body to test and validate AI applications, which directly informs local policy and future standards.

Werner praised Indian telecom operators distributing access to premium AI chatbots bundled with plans, stating it is a great way to increase adoption.
“But simply putting the tool in people’s hands is not enough,” he said. “Similar to mobile payments, unregulated and unguided access can lead in any direction.”The UN body has established an international AI standards exchange database to coordinate global efforts and prevent overlapping work with other bodies. The database currently hosts about 800 AI-related standards, Werner said.

“While governments and organisations frequently create broad declarations and governance frameworks, the ITU uses standards to actually implement these ideas,” Werner said, adding that the goal is to bake-in critical elements such as security, safety, privacy, human rights, and sustainability to AI technologies.

“We should ensure these are baked-in to AI technologies from the very beginning, rather than relying on a band-aid approach to fix problems after deployment,” Werner said.

However, the timing to implement standards is also important. If standards are implemented too early, they risk stifling innovation.

But if they are implemented too late, a few major players could dominate the market, leading to monopolies and vendor lock-ins.

The UN body is currently developing targeted standards to address major hurdles such as fighting misinformation, where ITU has fast-tracked standards on multimedia content authenticity to help verify if the media is original or altered. It is also preparing telecom networks for AI’s insatiable data and energy demands, where it is designing future 6G network specifications that can handle this bandwidth.