Who Is Really Benefiting From Free AI? — Beyond Tomorrow
Data residency is an important requirement in India's financial sector, thanks to zealous regulators but there's no such guardrails when it comes to GenAI.
There are no free lunches when it comes to apps and services. As the saying goes, when something is free, you are likely paying for it with your personal data. The most recent area where this could be true is the bonanza from AI companies for Indian users — a slew of premium AI services are now available at no cost for millions of Indians.
It started with Perplexity being offered free for a year via Airtel for its users. Then Google offered Gemini Pro free for students. Now Google is offering Gemini Pro free for millions of Jio users. And ChatGPT Go, a budget paid plan is free for all users in India for a year. And just today, fintech giant PhonePe has announced a collaboration with OpenAI where PhonePe users will be able to explore ChatGPT's features.
It's certainly understandable why the telcos are bundling these plans. Just like OTT bundling allowed them to offer a wider range of offerings to hook users and increase revenues, premium AI services could do the same. Even if a tiny fraction of users opt in for paid plans once the freebies run out, it's welcome revenue for telcos eager to increase ARPU (Average Revenue Per User) in a market where data rates are among the cheapest in the world and where the payoff from investments on advanced services like 5G is still in the hazy distance.
It's the AI companies that we should perhaps be more worried about. Given India's diversity—linguistically and otherwise, and the massive data usage on the part of users here, using premium AI tools will make AI usage stickier for more of us and our data will end up training these global GenAI models. Stories of horrified Indian parents discovering their kids have used an AI tool to do their homework do not even elicit a chuckle nowadays—it has become endemic. And we seem to be gifting away our data.
Greyhound Research describes the issue as a land grab, explaining, "These early interactions with AI are not neutral. They influence how users learn to ask questions, what formats they trust, and what types of responses they normalise. This is not just behavioural shaping. It is cognitive scaffolding. If your first few hundred AI interactions happen inside one platform’s model, that model becomes your epistemic lens. It does not just answer your questions. It starts to frame your worldview."
The government is certainly aware of the issue and needs to get moving quickly as Indians hand over what is often personal data to AI models. Another problem is at the enterprise level as free access to premium AI tools means employees will use them too. After all, children using AI to do their homework won't be the only ones playing truant. The consequences of confidential or client data being laid at the altar of GenAI tools in the quest for a shortcut to getting work done is the stuff of nightmares for most CIOs and CISOs. Especially when our data goes across the world and audit trails may not be possible too. Data residency is an important requirement in India's financial sector, thanks to zealous regulators but there's no such guardrails when it comes to GenAI.
I'm not playing doomsday prophet here. We cannot afford to lose ground in the AI race and it's already clear that India has a long way to go to catch up when it comes to models and AI infrastructure. Which is one of the reasons global AI leaders have the Indian market in a vise-like grip. There are no comparable Indian alternatives available. While the government is trying to catalyse change and supporting ambitious Indian attempts at building AI models, what this could means today though is that governance frameworks are urgently needed. This is especially true because India's Digital Personal Data Protection (DPDP) Act is still in the process of being implemented, though it was passed in 2023. This ball is certainly in the government's court.
Meanwhile, here are some of the most important AI-related reads from the past few days:
PhonePe Collaborates With OpenAI To Bolster ChatGPT Access In Big AI Push For Indian Users
GPT-5.1 Update Boosts ChatGPT's Brainpower And Personality Range — Check Details
Industry Seeks Narrower AI Labelling Rules, Warns Against Overreach
Whisper Leak Attack: Microsoft Warns About Unauthorised Access To Encrypted AI Chats By Hackers
Google Pixel Phones To Get AI-Powered Notification Summaries, Scam Alerts In Latest Update
SoftBank Sells Nvidia Stake For $5.8 Billion To Fund AI Bets
Meta Chief AI Scientist Yann LeCun Plans Exit To Kickstart New Startup
Intel’s AI Leader Sachin Katti Joins OpenAI, To Focus On Artificial General Intelligence
How Is Google’s Nano Banana 2 AI Photo Editor Better Than Nano Banana?
AI Will Take Over Human Jobs In A Decade: DeepSeek Researcher
AI With An Ego? Being Rude To ChatGPT May Improve Accuracy, Study Finds
'Potentially Catastrophic': OpenAI Issues Stern Warning On AI's 'Superintelligence'—Here's Why
Instagram Restyle Vs Google Gemini 2.5 Flash: Which AI Photo Editing Tool Is Better?
India’s AI Investments Rise, But Innovation, ROI Key Hurdles: Kyndryl Report
AI Boom Or Bubble? Big Tech's $400-Billion Gamble Echoes Dot-Com Era Fears
How AI Companies Are Hiring Small Town Workers To Train Robots
Tipping Point For AI Takeover? 'Humanoids At Cost of SUVs', Brett Adcock Tells Nikhil Kamath
Meet Hoxo — First AI-Powered Humanoid Robot Deployed In Nuclear Sector By Capgemini, Orano
Till next week,
- Ivor Soans
