Context:

• As Artificial Intelligence (AI) spreads into government services, recruitment and healthcare, and India courts $200 billion in AI investment over the next two years, disabled citizens are being brought into the conversation late, if at all.
• In this context, this article argues that AI's promise for disabled people will not reach them on its own — it requires deliberate design and policy choices.

A Real Example of AI's Promise

• A person, blind since birth, uses AI almost daily — to read documents, understand photographs, or navigate government forms his screen reader cannot open — all without relying on a sighted person.
• This independence in everyday tasks shows the technology's genuine potential. But this promise cannot be assumed to reach disabled Indians automatically.

The Legal Backdrop: A Right Already Recognised, But Unenforced

• November 2024 — Rajive Raturi judgment (Supreme Court)
  • The Court held that accessibility is a facet of the fundamental right to life and dignity.
  • It found India's existing accessibility rules were "toothless" — merely polite suggestions, when the Rights of Persons with Disabilities Act, 2016 intended binding standards.
  • The Court gave the Centre three months to frame mandatory standards.
• Nearly Two Years Later: The petitioners are back in court, because too little has changed.
• Analysts argue that the disabled people should not be made to adapt to systems built without them — and AI is simply the newest such system.
• Enforcement Failure on the Ground
  • The Chief Commissioner for Persons with Disabilities penalised 155 establishments, including government ministries, for websites and apps a disabled citizen simply cannot use.
  • Though the 2016 Act has required accessible digital services since 2019, compliance remains the exception, not the rule.
  • This is the ground reality on which AI is now being deployed.

AI's Inherent Bias Against Disability

• AI is not neutral — research shows it performs systematically worse when disability enters the picture.
• When researchers ran 21 language models through it, the models became more error-prone, more negative in tone, and more likely to stereotype the moment disability entered the question.
• Everyday Manifestation: Tell a mainstream chatbot you are blind and want to become a software engineer, and it often opens with "I'm sorry you're blind." The person asked a career question; the model responded as if they had reported a loss.

The Infrastructure Blind Spot: Data Centres and Power

• The neglect extends beyond software bias into physical infrastructure.
• India's data-centre capacity is projected to more than quadruple by 2030, reaching 6.5 gigawatts or more.
• States are competing for these projects through power subsidies and duty waivers.
  • For example, Maharashtra this year relaxed its renewable-energy requirement for data centres from 100% to 51%.
• Rarely does anyone ask: what does this new load do to a grid already straining at summer peaks, or what does server heat add to cities already dangerously hot?

Why This Matters Specifically for Disabled People?

• This is not a distant environmental debate for millions of disabled Indians — it is immediate and personal.
• Many depend on powered wheelchairs, oxygen machines, or other equipment that cannot be switched off when the grid strains.
• Disabled people are often the first to be stranded when a power cut or an inaccessible emergency alert disrupts daily life.

Way Forward

• Mandatory bias testing — any AI system the government deploys should be tested for disability bias.
• Genuine representation in training data — datasets need authentic disability representation, gathered with consent.
• Conditional state incentives — incentives for data centres should carry conditions on renewable sourcing and on grid and heat impact.

Conclusion

• This is not an argument against AI or against the infrastructure a growing economy needs.
• However, disabled people should not be asked to surrender tools they depend on in the name of sustainability. They must have a voice in shaping this technology — from the data that trains the models to the grids that keep them running.