Airtel and Jio’s AI push offers big promise—and bigger risks

Bharti Airtel was the first off the blocks, offering a year’s free subscription to the Pro model of Perplexity AI to its 390 million subscribers across mobile, wi-fi and DTH broadband.

Reliance Jio followed suit, partnering with Google to offer a package comprising 18 months of free subscription to its advanced AI model Gemini Pro, bundled with 2TB of cloud storage, plus access to its suite of AI-driven content creation tools. While Jio has a subscriber base of over 505 million, not all of them will be eligible to get the Gemini add-on.

Initially, Jio is targeting its premium plan broadband subscribers in the 18-25 age group, with a national rollout planned for a later stage.

Though disaggregated age-wise subscriber data is not available, given that 40% of the population is under 25 and 43% under 35 years of age, the potential user base across the two telecom operators rises into the hundreds of millions, especially since OpenAI has also made its advanced ChatGPT model free for Indian users for a year.

That will put the Indian AI user market on par with or ahead of China in terms of AI user base, and significantly ahead of the US and Europe.

It can cause a paradigm shift in how global AI majors view and approach the Indian market. According to a report by resourcera.com, the global market for generative AI chatbots alone was valued at $7.66 billion in 2024 and is expected to reach $65.94 billion by 2032, with a CAGR of 31.1% over the forecast period.

However, that figure was estimated before the telco-AI deals happened in India, which have the potential to dramatically alter the user base, and thereby the market size.

AI for millions

It is clear why global AI majors are making a play for India. With the world’s largest number of mobile telephone users and internet users outside China – a market they cannot access – India stands out as one with the potential for future growth. Bundling with a telco slashes customer acquisition costs, while enabling near-instantaneous national-scale rollouts.

It is also clear why telcos are making this play. If video and OTT drove the uptick in data consumption, paving the way from 3G to 4G to 5G, AI – and the productivity tools that they come bundled with – are seen as the next growth engine for telecom players in their effort to convert their massive user bases into higher average revenue per user.

What is not clear is the impact this near-universalisation of access to AI tools will have on the country as a whole. While the benefits are unquestionable—improving learning outcomes, enhancing productivity, and opening up the market for a host of other services, particularly in agriculture and small businesses—the risks are also very real.

India’s huge base of internet users is largely comprised of newcomers to the web, many of whom are relatively inexperienced or unaware of the risks. This is evidenced by the rising tide of digital scams and fraud proliferating in India. Add AI to the mix and the risks jump manifold. Nobody has estimated the impact of population-scale exposures to AI errors, hallucinations, and deepfakes.

Ensuring this does not happen will require the active intervention of all stakeholders, including government and regulators, AI intermediaries, and telecom service providers.

Guardrails must be built into AI systems — tweaking how AI results are displayed, displaying citations and sources by default, providing information about the degree of confidence one can place on the results, as well as labelling content as envisaged under the draft IT rules amendments.

AI platforms must also incorporate India-specific features for young users and provide clear disclosures and warnings related to finance, health, and other relevant areas. An alarming number of people already research symptoms using AI, and in a country where there is little regulation over the sale of prescription medicines, this can lead to serious consequences.

All stakeholders also need to come together to develop and implement easy-to-access educational programs that can help users understand the advantages and risks of using AI tools.

A grievance redressal mechanism should be developed to define the roles and responsibilities of different parties – AI service providers and telecom service providers.

India’s policymakers may have been caught off guard by the speed of these developments, but they cannot afford any more delay in getting their act together.