The Tech Law Paradox: Delhi HC Shields OpenAI with "Fair Dealing" while Supreme Court Grapples with DPDP's RTI Blindspot

The Twin Pillars of India’s Data Economy are in Judicial Limbo

For Indian technology and intellectual property lawyers, 2026 is shaping up to be a year of profound cognitive dissonance. On one hand, courts are bending over backward to stretch mid-20th-century copyright defenses to protect 21st-century artificial intelligence. On the other, the foundational privacy framework—the Digital Personal Data Protection (DPDP) Act, 2023—remains mired in constitutional challenges and bureaucratic inertia.

Two major developments this week perfectly encapsulate this paradox. The Supreme Court has referred the formidable challenge against the DPDP Act’s controversial amendment to the Right to Information (RTI) Act to a larger bench, crucially without granting a stay. Meanwhile, the Delhi High Court has delivered a blockbuster prima facie ruling in OpenAI v. ANI, arguably reshaping Indian copyright law by squeezing Large Language Model (LLM) training into the narrow confines of "fair dealing."

The AI Copyright Bombshell: Is Commercial LLM Training "Private Research"?

Let’s start with the IP earthquake. In the ongoing saga of OpenAI v. ANI, the Delhi High Court has taken a controversial preliminary stance. The Court held prima facie that the storage and ingestion of ANI’s literary works (news reports, articles) for the purpose of training OpenAI’s LLMs could fall squarely within Section 52(1)(a) of the Copyright Act, 1957.

For the uninitiated, Section 52(1)(a) exempts "fair dealing with any work" for the purposes of "private or personal use, including research."

This is a staggering judicial leap. To categorize the mass scraping of proprietary data by a multi-billion-dollar commercial enterprise as "private or personal use" or mere "research" stretches the statutory language to its absolute breaking point. It fundamentally ignores the commercial realities of generative AI.

"By classifying LLM training as 'private research', the Delhi High Court is effectively subsidizing Silicon Valley's AI race with the intellectual property of Indian content creators."

Why this matters for your practice: If you are representing publishers, media houses, or digital creators, your cease-and-desist letters to AI scrapers just lost a lot of their bite. Until this is challenged or clarified by a Division Bench or the Supreme Court, tech companies will use this prima facie ruling as a carte blanche to ingest Indian copyrighted material. IP litigators must now pivot their strategy: you cannot rely solely on the act of reproduction; you must prove market substitution and commercial detriment to defeat the "fair dealing" defense.

The Supreme Court, DPDP, and the Death of the "Public Interest" Test

While the Delhi High Court focuses on proprietary data, the Supreme Court is wrestling with personal data. The apex court has referred challenges to both the DPDP Act, 2023 and the subsequent 2025 Rules to a larger bench. Notably, the Court declined to stay the operation of the Act.

The crux of the constitutional challenge lies in Section 44(3) of the DPDP Act, which amends Section 8(1)(j) of the RTI Act, 2005. Previously, the RTI Act allowed public information officers to disclose personal information if they determined that the larger public interest justified the disclosure. The DPDP Act brutally amputated this balancing test, creating a blanket ban on the disclosure of personal information under the RTI framework.

The petitioners rightly argue this creates an impenetrable shield for corrupt public officials. The Supreme Court is now tasked with defining the elusive boundary between "public data" and "private data"—a boundary that dictates platform regulation, media rights, and the very survival of transparency journalism in India.

What the lack of a stay means for practice: Because the Supreme Court did not stay the DPDP Act, corporate compliance mandates remain in full force. Following the implementation of the 2025 privacy regulations last November, your clients must continue overhauling their data collection architectures to ensure data minimization, clearer user notices, and robust opt-out rights. You cannot advise clients to "wait and see" what the larger bench decides; non-compliance is not an option.

The Phantom Data Protection Board

Adding a layer of dark comedy to the compliance scramble is the state of the Data Protection Board of India (DPBI). The 2025 Rules formally established the Board in law, yet as of mid-2026, the Chairperson and Member seats remain empty. Nomination communications were only issued in May and June of 2026.

This creates a bizarre regulatory vacuum. We have an active, stringent data privacy law (as affirmed by the SC’s recent ruling that student data under the APAAR Scheme is strictly subject to the DPDP Act and cannot be shared with private entities unlawfully), but the primary adjudicatory body exists only on paper.

The Strategic Takeaway: Do not mistake the lack of an active Board for a lack of legal risk. Data fiduciaries are currently operating in a high-risk environment where DPDP violations could be weaponized in consumer courts, writ petitions, or even by sectoral regulators like the RBI and SEBI who aren't waiting for the DPBI to wake up. Furthermore, as legal commentary increasingly highlights the overlap between DPDP compliance and workplace technology (like AI monitoring of employee data), employment lawyers must urgently integrate data privacy into HR policies.

The Verdict

Indian technology law is currently a landscape of aggressive judicial interpretation and sluggish executive implementation. The Delhi High Court’s willingness to shield AI under Section 52(1)(a) shows a judiciary eager to foster innovation, perhaps at the expense of traditional IP rights. Conversely, the Supreme Court’s cautious handling of the DPDP-RTI conflict highlights the profound constitutional anxieties surrounding state control of information.

For the practicing lawyer, the mandate is clear: prepare for heavy litigation on the parameters of "fair dealing" in AI, and enforce strict, conservative DPDP compliance for your corporate clients, regardless of whether the regulatory watchdog has grown its teeth yet.