AI, equity & the next tech-divide. And why access to it, must become universal.

Let's dive into this pressing issue of our times - AI and the equity challenges it brings!

First, I look into the "what does it mean" part of this provocation in the headline.

Part 1 – The Provocation: AI isn’t a luxury. It’s a utility. And on the path to become a right.

We aren’t far away from AI access being debated as a basic human right.

Not a perk. Not a “nice-to-have.” Not just a productivity hack. But a utility - as foundational as electricity, internet, or education itself.

And if we fail to frame it this way, we risk deepening divides:

• Between rich and poor nations.
• And even more dangerously, within poorer nations themselves.

It is because the knowledge, insight, and action amplification that AI platforms bring is massive - and increasingly decisive.

• In 2016, the UN declared internet access a human right, because lack of it meant exclusion from education, jobs, and government services.
• Premium AI subscriptions today (ChatGPT Pro, Gemini Advanced, Claude, Perplexity Pro et al) deliver 10x efficiency jumps. In my own sustainability and circularity work, what would take 20 days to produce, can now be done in half a day and often with much better insights and pattern recognitions.
• Farmers use AI-driven SMS tools to diagnose crop diseases in real time. Without it, their peers wait weeks for an extension officer, risking entire harvests. This is happening across Kenya, Brazil and India.

Hence, the gap doesn’t just widen - it compounds. The “AI-haves” and the “AI-have-nots” are on diverging trajectories.

Hence my provocation is that AI is not a luxury anymore. It’s on the clear path to becoming an universal need.

Part 2 – Addressing the misleading trade-offs on "Do we spend on 'education, food, water, health' or AI?"

Whenever I raise this, the most common pushback comes in the form of trade-offs.

“India must first focus on education, food, water, and healthcare before AI.” “If India had to choose between AI and primary education, it’s obvious where the priority lies.”

But these self-imposed hierarchies can be mis-leading, as AI doesn’t compete with basic rights - it multiplies our ability to deliver them. Let’s look at this, through a few clear examples.

Education

In the last two years of teaching climate risk classes, I’ve seen this first-hand. Students now run AI chats alongside lectures. Trainers who come underprepared, rely on rote notes, or deliver robotic sessions, are exposed instantly.

That forced me to level up. I now prepare with 5x more effort, bringing sharper insights and lived-experience examples. In fact, being a trainer, I am able to ask much more layered questions to an AI platform, go further down rabbit holes of prompts and conversations, to get to an even better understanding – for myself as well as a student just typing away one quick query on it. AI didn’t undermine me - it raised the bar for me, and my own preparedness, and most certainly for the classes I took in general.

Ed-tech platforms show the same shift:

• Khan Academy’s Khanmigo acts as a personalized tutor.
• Byju’s in India is experimenting with AI tutors for millions of students.
• Students prepping for UPSC and CAT use AI to build structured study plans, simulate tests, and get feedback.

This isn’t AI vs education. It’s AI for education.

Food, water, and healthcare

The same holds true for basic needs.

We now know of umpteen cases of when quick AI sessions gave so many of us so much more clarity on our parents’ health records than multiple hospital visits.

By using any combinations of Claude, Gemini, ChatGPT, Perplexity etc; multiple manufacturing companies, consulting firms, not-for-profit bodies are now getting prompt access to insights from 50+ reports at a single time – helping reduce time to grasp that research, plus getting much more inter-connected pattern recognition and also getting our solution models stress tested with possible success or failures of the same from across the world.

And from what I read in a recent news report, Apollo Hospitals has now cut radiology scan wait times by 60% with AI. Across agriculture, too, AI is enabling pest detection, yield optimization, and climate-adaptive practices - something which becomes even better when aspects of historical AI analytics is triangulated with drone and satellite mapping!

So why do we frame AI as an access tool that comes “after” the basic rights of food, water, and health? It’s already helping deliver them more efficiently and effectively. The ones ignoring it for its hallucinations, and hence not using it, will be late to realise that simple prompts and research repository creations can help address it anyway.

Access vs Capability

The other trade-off argument is: “Let’s first build domestic AI capability and then expand consumer access.”

But this sequencing is flawed. Look at India’s Jio 4G revolution:

• We didn’t wait for domestic phone manufacturing before enabling cheap internet access.
• Access exploded first. Demand then drove Make in India.

If we hold back AI access until “domestic capability” is built, we risk having neither robust access nor strong capability. Worse, we risk losing an entire generation of students, workers, and entrepreneurs who could be leveraging AI today. Imagine if we first insisted on our capability on computer hardware and software, to be able to leverage its end-use? We would still be stuck in the 1990’s.

Quite clearly - access to tech/AI itself will enable capability, and not the other way around. 

Part 3 – The equity challenge: The winners, losers and how we can prevent a further divide?

Here’s where the debate hits hardest: who gets squeezed if AI access stays gated by its current disposable income filter?

The obvious answer is “the poor.” But the real immediate losers might be the middle class. And that’s because at the top, the elites and corporates can afford premium AI, so they accelerate. At the bottom; eventually, government or NGO programs may provide AI touchpoints. But in the middle: professionals without disposable income for AI subscriptions fall behind their peers - and behind the technology curve itself.

The middle-class is getting squeezed

We see this every day across our work lives

• With premium AI access, many senior leaders now outperform entire entry-level teams. The work of 20 days? Done in half a day. Those entry-level and even mid-level roles vanish – We are seeing the immediate and consequential impact of it already in thousands of IT roles in Indian IT companies.
• An environment lawyer with GPT-4 Pro drafts briefs 5x faster. His equally skilled peer without AI (the one who is anyway behind the generational wealth race) loses billable hours and eventually clients.
• UPSC and CAT aspirants with AI generate structured prep strategies and instant feedback. Their peers without AI are stuck with older methods, and heavily dependent on older irrelevant methods.

This isn’t the digital divide we solved with internet access. This is Digital Divide 2.0 - about who has AI leverage and who doesn’t.

Solutions would require the policy apparatus to proactively prepare for this, as

We’ve tackled equity challenges before:

• Right to Education
• Right to Food
• Right to Internet

A Right to AI Access in a country like India could look like:

1.      Subsidized AI credits -> Bundled into education, healthcare, and MSME schemes (which can be used by those preparing for competitive exams, frontline health workers, and micro-entrepreneurs).

2.      AI kiosks in rural India -> Built on CSC network for translations, subsidy claims, agriculture advisory, and healthcare support. These can be available at the Panchayat offices, and schools across rural India – no new hardware or infra upgrade required.

3.      Baseline public AI utility -> Govt-backed, multilingual, open-source AI for everyday use (summaries, translations, forms, diagnostics) can do wonders - and almost becomes the base of AI as a universal access tool.

4.      Public-private AI rails (like the UPI model) -> Govt sets open APIs/datasets, startups and corporates innovate on top, ensuring universal access. What the team of BeckN Protocol is doing, has a lot of learnings for everyone.

5.      AI integrated into welfare schemes -> Integrated into PM-Kisan, MGNREGA, Ayushman Bharat for benefits, claims, and grievance redressal – imagine farmers chatting with a database-rich AI for crop advisory, patients using AI for getting a quick reading of their latest reports. And imagine this now in local languages?

6.      Regional Language Equity -> Right to AI Access must guarantee vernacular parity: AI that speaks in Marathi, Bhojpuri, Khasi, or Manipuri, not just English/Hindi. Govt could fund datasets & benchmarks through Bhashini (India’s language AI mission). This was just how Doordarshan’s regional channels made governance inclusive.

7.      AI for Skilling & Jobs -> Just like railways gave millions access to mobility, AI can give millions access to employability. AI also becomes a digital logistical enabler for the youth to appropriately up-skill and find business/job opportunities, without the chained need for exploitative intermediaries. It gives even more agency to our youth.

8.      AI-Enabled Public Libraries -> We may revamp India’s public library network (Nehru Memorial Libraries, state libraries) into AI Learning Hubs. Like free Wi-Fi at railway stations, public AI at libraries can be an on-ramp to opportunity. 

Plus, the need for balancing regulation and innovation is more important than ever before

One last false binary that needs to be addressed is the fight of regulation vs innovation.

• The EU AI Act is already criticized for stifling startups compared to the US.
• Over-regulation too early risks “innovation deserts.”

But… no regulation risks bias, misinformation, deepfakes, and exploitation.

India has the chance to do what it did with UPI:

• Build public-good infrastructure at scale.
• Allow private players to innovate on top of it.
• Ensure access first, regulate sensibly later.

To summarise, AI isn’t in competition with basic rights - it’s already multiplying them.

The real debate is not whether AI should come after food, water, education, and health - but how to make sure AI is equitably accessible, so it enhances those very rights for everyone, not just the few.

The choice is clear:

• If we frame AI as a luxury, we widen inequality.
• But, if we frame AI as a universal right, we close divides and unlock yet unknown opportunity.

So, should AI access be treated as a basic human right? Or will we wait until the divide becomes unbridgeable?

PS - Of course, I used multiple AI subscription models (Perplexity Pro, Gemini Pro, GPT Premium) to help me categorize and fine-tune my original write-up into these three parts!