The Lift Line
Most of India’s technology law was written for a world where humans clicked and machines merely obeyed. Agentic AI that acts on its own and deepfakes that manufacture reality break that assumption. The question before policymakers is no longer whether to regulate artificial intelligence, but how to assign accountability when the actor is a piece of software.
Why This Editorial Matters for Your Exam
Regulation of emerging technology is one of the most reliably examined themes because it forces candidates to reason about governance keeping pace with innovation. AI is the frontier case, touching rights, markets and national security at once.
The topic also tests whether you can move past slogans about ethics to the concrete design of a statute, distinguishing what the law should mandate from what industry codes or standards should handle.
GS Paper 2: Government policies and interventions for development in various sectors, and the regulation of new technologies and their governance.
GS Paper 3: Awareness in the field of IT, and the challenges of cyberspace, artificial intelligence and content authentication.
For Prelims, hold the specifics: India currently regulates online content through the Information Technology Act, 2000 and the IT Rules, 2021 on intermediary due diligence, with data protection under the Digital Personal Data Protection (DPDP) Act, 2023; the IndiaAI Mission anchors capacity-building and advisory work on AI governance; the EU AI Act is the leading example of a risk-based statute; and the core technical challenges are agentic AI, autonomous systems that take actions, and deepfakes, AI-generated synthetic media that require watermarking and labelling.
For Mains, the demand is to argue for a specific regulatory design and defend it against both over-regulation and under-regulation.
Background and Context
For over two decades India’s digital governance rested on the IT Act, 2000, built around the idea of an intermediary that hosts content created by identifiable users. The IT Rules, 2021 added due-diligence duties, and the DPDP Act, 2023 created a data-protection regime. None of these was designed for systems that generate content autonomously or take actions in the world without a human in the loop.
Two developments have exposed the gap. First, deepfakes now let anyone fabricate convincing audio and video, weaponising synthetic media for fraud, defamation and disinformation. Second, agentic AI systems do not merely answer questions but execute tasks, book, transact, and chain actions together, raising the question of who is responsible when they cause harm.
India has so far preferred a light-touch, pro-innovation posture, using advisories and the IndiaAI Mission’s committee work rather than a hard statute. The debate now is whether that posture is sufficient, or whether a dedicated, calibrated law is needed.
The Core Argument / Issue
The attribution problem
When an autonomous agent errs, liability is genuinely hard to locate. Is it the developer of the model, the deployer who fine-tuned it, the platform that hosted it, or the user who set the goal? A modern AI law must set default rules of accountability across this chain rather than leaving courts to improvise.
Authenticating reality
Deepfakes attack the shared trust that democracy and commerce depend on. The regulatory answer lies less in banning generation than in mandating provenance, watermarking and clear labelling of synthetic media, so audiences can tell the genuine from the generated.
Risk-based, not blanket
A single rule for all AI is either too heavy for low-risk uses or too light for high-risk ones. The sensible design tiers obligations by risk, as the EU model attempts.
| Approach | Core idea | Trade-off |
|---|---|---|
| Light-touch advisories | Voluntary codes, minimal statute | Fast to innovate, weak on accountability |
| Risk-based statute | Duties scale with risk of the use case | Balanced but complex to administer |
| Blanket hard regulation | Uniform strict rules for all AI | Strong protection, risk of stifling startups |
How to Think About This (Analytical Frame)
Think in terms of calibration, not prohibition. The goal is a principles-based law that fixes clear accountability and provenance duties while leaving technical detail to evolving standards, so the statute does not become obsolete with the next model release. Weigh two failure modes. Over-regulation drives talent and startups abroad and cedes the field to larger economies. Under-regulation lets harms accumulate until trust collapses and a harsher clampdown becomes politically inevitable. The wise path anchors a few durable obligations, transparency, accountability, human oversight for high-risk uses, and lets regulators and courts fill in the rest. This reflects India’s interest in remaining a leading, sovereign AI power while protecting citizens.
The Diagram in Words
Autonomous AI agents and deepfakes -> gaps in IT Act and IT Rules -> attribution and trust problems -> risk-based principles law -> clear accountability plus content provenance -> innovation with safeguards
Way Forward
- Legislate accountability. Enact a calibrated AI law that assigns default liability across the developer, deployer and user chain for autonomous agents.
- Mandate provenance. Require watermarking, labelling and traceability of synthetic media so deepfakes can be identified and misuse penalised.
- Tier by risk. Adopt a risk-based structure, with strict duties for high-risk uses and a light touch for low-risk innovation, drawing lessons from the EU AI Act.
- Build capacity. Empower a competent regulator through the IndiaAI Mission, invest in auditing and testing, and keep rules principles-based so they age well.
PYQ Linkage and Practice
The UPSC has asked about the impact of digital technology on governance and about cyberspace challenges, for example the 2020 question on cyber security. An AI-regulation question is a natural extension into the frontier of autonomous systems and synthetic media.
Practice question: “India needs an AI law that fixes accountability for autonomous agents and deepfakes without stifling innovation.” Discuss the gaps in the existing framework and the design of a calibrated AI regulation. (15 marks, 250 words)
Sources: Indian Express, Designing India’s AI law, MeitY, IndiaAI Mission
Source: Writing India's AI Law for Autonomous Agents and Deepfakes — Ujiyari.com | Free UPSC & State PCS Editorial Analysis
