The DPDP Act's Trial by Fire: Constitution Bench to Hear RTI Clash While Government Punts on AI Regulation

The Collision Course: Privacy, Transparency, and the AI Vacuum

Welcome to the June 2026 landscape of Indian technology law, where the dust has finally settled on the drafting of the Digital Personal Data Protection (DPDP) Rules, 2025, only to be kicked up entirely by constitutional challenges and regulatory complacency. For technology, IP, and corporate lawyers, this week’s developments present a chaotic but critical mandate: you must navigate a compressed compliance timeline for a law whose very constitutional validity is now sub judice before a five-judge Constitution Bench.

At the heart of the current storm are two distinct but equally consequential developments. First, the Supreme Court has escalated the constitutional challenge against the DPDP Act to a larger bench while flatly refusing an interim stay. Second, the Ministry of Electronics and IT (MeitY) has officially signaled that India will not see bespoke Artificial Intelligence legislation anytime soon, relying instead on the existing DPDP and Intellectual Property frameworks to do the heavy lifting. Both developments demand immediate strategic pivots from practicing lawyers.

The RTI vs. DPDP Showdown: Why the Supreme Court Escalation Matters

The Supreme Court’s decision to refer a clutch of petitions—led by the Editors Guild of India, The Reporters' Collective, and prominent RTI activists—to a five-judge Constitution Bench is a defining moment for Indian information law. The petitioners have zeroed in on the fatal flaw of the DPDP Act: its amendment to the Right to Information (RTI) Act, 2005.

Prior to the DPDP Act, Section 8(1)(j) of the RTI Act provided a nuanced exemption for personal information, allowing Public Information Officers (PIOs) to disclose personal data if the "larger public interest justifies the disclosure." The DPDP Act obliterated this balancing test. By amending Section 8(1)(j) to provide a blanket exemption for any information relating to "personal data," the legislature effectively weaponized privacy against public accountability.

For litigators and activists, this is a textbook constitutional clash between two facets of Article 19(1)(a)—the right to know versus the right to privacy (as recognized in Puttaswamy). However, Chief Justice Surya Kant, alongside Justices Joymalya Bagchi and Vipul M. Pancholi, made the immediate operational reality brutally clear:

"There is no question of stay. We will not thwart the regime introduced by Parliament until the case is heard and a constitutional determination is made."

The Practice Takeaway: In-house counsels and compliance officers cannot use the pending Constitution Bench reference as an excuse to delay DPDP compliance. With MeitY hinting at compressing the 18-month transition period for large data fiduciaries (originally set to end in May 2027), the compliance runway is vanishing. You must proceed with overhauling privacy notices, establishing consent architectures, and instituting breach protocols immediately. A bet on a judicial stay is a losing wager.

The "Existing Laws" Fallacy: Shoehorning AI into the Copyright Act

While the Supreme Court tackles the RTI dilemma, the government has dropped a regulatory bombshell regarding Artificial Intelligence. MeitY Secretary S. Krishnan recently articulated the government’s stance: there is no need for new AI laws, as the DPDP Act and the existing Intellectual Property (IP) framework cover a "fair amount of issues."

From a practitioner's standpoint, this "wait and watch" approach is deeply problematic. Expecting the Copyright Act, 1957, and the nascent DPDP Act to comprehensively govern generative AI is like expecting a bicycle mechanic to service a jet engine. Let's examine the cracks in this foundation.

Under Section 12 of the DPDP Act, data principals have the explicit right to seek correction, updating, and erasure of personal data. How does this apply to Large Language Models (LLMs)? If a data principal's personal information is ingested into an AI model's training data, standard deletion protocols fail. "Machine unlearning"—the process of making an AI model forget specific data without retraining the entire model from scratch—is technologically nascent and financially exorbitant. When your client receives a Section 12 erasure request targeting an AI model, arguing "technical infeasibility" will not fly under the current strict liability framework of the DPDP Act.

Furthermore, relying on the current IP framework to govern AI training data scraping will inevitably lead to a flood of litigation. Does ingesting copyrighted works to train an AI model fall under the "fair dealing" exception of Section 52(1)(a) of the Copyright Act? Indian courts have traditionally interpreted fair dealing restrictively. Without bespoke AI legislation clarifying text and data mining (TDM) exceptions, IP lawyers will be left fighting protracted battles of first impression in the High Courts, relying on disparate foreign jurisprudence from the US and EU.

The Road Ahead for Practitioners

The intersection of DPDP compliance, constitutional uncertainty, and a deliberate regulatory vacuum for AI means that Indian tech lawyers are currently operating without a safety net.

To advise your clients effectively this quarter, you must:

1. Accelerate DPDP Audits: Treat the 18-month transition period as a suggestion, not a guarantee. The government’s intent to compress this timeline for major tech players means enforcement could begin sooner than anticipated.
2. Draft AI-Specific Data Policies: Since the government is relying on the DPDP Act to govern AI, your data processing agreements must explicitly account for how third-party AI vendors utilize your client's data. Ensure indemnities cover failures in "machine unlearning" related to Section 12 erasure requests.
3. Prepare for IP Litigation: If you represent creators or media houses, prepare to test the boundaries of copyright infringement against AI scrapers. If you represent AI startups, you must build robust fair-use defenses now, as statutory safe harbors are not coming to save you.

2026 is shaping up to be the year where theory meets practice in Indian technology law. The laws on the books are being stretched to their absolute limits—and it will be up to the legal practitioners to hold the seams together.