India is entering its most consequential technology moment since the internet went mainstream. Artificial intelligence promises productivity gains, new industries, and global leadership. But it also brings a harder question: How do we innovate at scale without eroding human dignity?
The recent controversy around Elon Musk-owned Grok, blocked by Malaysia and Indonesia over the misuse of non-consensual deepfakes, should be read in India not as a foreign regulatory skirmish, but as an early warning. It reveals what happens when powerful generative tools are deployed faster than the safeguards needed to protect citizens.
The lesson for India is clear. AI governance is no longer a future concern. It is a present responsibility.
When technology scales harm
Generative AI has collapsed the cost of creation. Text, images, audio, and video can now be produced instantly, at near-zero marginal cost. That same power also collapses the cost of harm.
Non-consensual sexual deepfakes are not edge cases. They are a predictable outcome when realistic image and video generation is made frictionless and widely accessible. The damage is not abstract: Reputations destroyed, consent erased, trauma multiplied, often with little recourse for victims.
This is why Indonesia described such content as a violation of human rights and dignity, and why Malaysia criticised platforms for relying on reactive reporting rather than proactive controls.
These are not moralistic objections. They are governance responses to a technology that scales faster than social repair mechanisms.
India will face the same pressures and at a far greater scale.
Also read: India’s ancient knowledge could save the future of AI. Build with wisdom, not just engineering
India cannot afford a binary debate
India’s instinct has often been to frame technology debates as innovation versus regulation. That framing is outdated.
The real choice is between trust and backlash.
History is instructive. Industries that fail to self-correct invite blunt regulation later. Social media learned this after years of ignoring misinformation and abuse. AI is approaching a similar inflection point.
India, however, has a unique advantage. It has already built digital public infrastructure such as Aadhaar, UPI, DigiLocker that combines scale with safeguards. The success of these systems did not come from the absence of rules, but from thoughtful design, legal oversight, and iterative governance.
AI should follow the same path.
Human dignity must be a design constraint
If there is one principle India should make non-negotiable, it is this: Human dignity is not optional in the age of AI.
Free expression does not include the right to algorithmically impersonate, sexualise, or humiliate real individuals without consent. When AI systems can foreseeably enable such harm, platforms cannot hide behind neutrality.
India’s Ministry of Electronics and Information Technology has already signalled that safe harbour protections may be reconsidered if AI-generated harm is not addressed. This is a necessary evolution. Safe harbour was designed for passive intermediaries, not for systems that actively generate content.
Clear expectations are pro-innovation, not anti-innovation. Uncertainty is what scares founders and investors, not rules.
Also read: India is letting the AI revolution bypass the country. It may have to pay a heavy price
Responsibility should match capability
India should adopt a simple, pragmatic standard: The more powerful the capability, the higher the responsibility.
This does not mean criminalising developers or stifling research. It means requiring reasonable, proactive safeguards where risks are obvious. Content controls, friction for high-risk use cases, auditability, and grievance redress mechanisms are not ideological demands. They are basic hygiene for mass-scale systems.
Importantly, this approach preserves space for startups to experiment, while drawing firm lines around deployment that affects millions.
Innovation and protection aren’t opposites
India’s ambition to become a global AI hub is legitimate and necessary. We need more compute, more talent, more open models, and more applied AI startups solving real problems in health, education, manufacturing, and governance.
But innovation divorced from social trust does not endure.
The platforms that last are not those that push boundaries the fastest, but those that align machine power with human values. Trust compounds. Distrust explodes.
By articulating a balanced framework of innovation with dignity and openness with safeguards, India can avoid the whiplash seen elsewhere, where public outrage forces sudden, heavy-handed regulation.
India’s global role
With its population size, democratic structure, and digital scale, India is not just another market.
It is a rule-maker by default.
If India demonstrates that it is possible to govern AI responsibly without killing innovation, it sets a template for the Global South i.e. countries that want growth without importing either Silicon Valley’s excesses or Europe’s regulatory complexity wholesale.
This is a strategic opportunity.
Also read: AI is learning caste bias in India. Who will audit it for discrimination?
The way forward
India should not wait for crisis-driven regulation. It should lead with clarity.
- Treat human dignity as a hard constraint in AI deployment.
- Align responsibility with capability, especially for mass-scale systems.
- Encourage innovation but insist on safeguards where harm is foreseeable.
- Build governance iteratively, as we have done with digital public infrastructure.
AI is writing the first draft of the future. India has both the scale and the institutional experience to help edit it wisely.
If we get this right, we won’t just build competitive AI companies. We will build a digital society worthy of the trust of a billion citizens.
And that is innovation in its truest sense.
Manish Maheshwari is General Partner at BAT VC, a New York-headquartered venture capital firm backing AI, Fintech and Enterprise SaaS startups. He previously led Twitter in India as Country Director. Views are personal.
(Edited by Theres Sudeep)