[Answered] Examine how the Delhi High Court’s ruling in ANI v. OpenAI balances technological innovation with intellectual property rights through judicial interpretation of fair dealing. (500 Words)

Introduction

Economic Survey 2025–26 identifies AI as a productivity multiplier, while Budget 2026–27 prioritises the IndiaAI Mission. Against this backdrop, ANI v. OpenAI redefines copyright jurisprudence by applying technologically neutral fair dealing principles.

Delhi High Court’s Judicial Interpretation of Fair Dealing

1. Technology-Neutral Interpretation of Research: Interpreted Section 52(1)(a), Copyright Act, 1957 dynamically through the doctrine of updating construction. Extended “research” beyond human activity to machine-mediated learning, preserving legislative intent. Prevents law from becoming obsolete amid technological evolution. Example: LLM training.
2. AI Training as Non-Expressive Use: Distinguished pattern learning from expressive reproduction. AI analyses statistical relationships rather than publishing original copyrighted expression. Mere ingestion of publicly available data is not infringement per se. Example: Token prediction.
3. Memorisation Test for Copyright Violation: Court held infringement arises only if AI systematically memorises or reproduces copyrighted content. Plaintiffs must establish actual output similarity instead of alleging data scraping alone. Raises evidentiary threshold while protecting legitimate copyright interests. Example: Verbatim outputs.
4. Fair Dealing as Public Interest Doctrine: Fair dealing is an internal balancing mechanism, not merely an exception. Injunction against AI training could impede India’s AI ecosystem, research institutions and startups. Balance of convenience favoured innovation at the interim stage. Example: IndiaAI Mission.

How the Judgment Balances Innovation with Intellectual Property

1. Indigenous Foundation Models: Lowers entry barriers for Indian AI startups by avoiding costly licensing obligations. Encourages indigenous foundation model development. Supports Digital Public Infrastructure-led AI innovation. Example: BharatGen.
2. Technology-Neutral Constitutional Dynamism: Reinforces technology-neutral interpretation consistent with constitutional dynamism. Aligns with Supreme Court’s preference against expansive interim injunctions in IP disputes (Bajaj Auto v. TVS). Preserves judicial flexibility pending final trial. Example: Interim jurisprudence.
3. Democratizing AI Innovation: Supports India’s aspiration of becoming a global AI hub under the IndiaAI Mission. Avoids concentration of AI development among capital-rich firms alone. Encourages startup competitiveness. Example: Deep-tech ecosystem.
4. AI-augmented Creativity: According to the London School of Economics (2025), AI is complementing rather than replacing many creative professions through specialised tools. Adobe (2025) reported over 83% creators already incorporate AI into creative workflows. Copyright should facilitate—not obstruct—new forms of creativity. Example: Creative AI.
5. Copyright Law: Harmonises Article 19(1)(a) (freedom of expression), Article 19(1)(g) (profession/business), and Article 300A (property rights). Balances creators’ proprietary interests with society’s interest in knowledge creation. Example: Constitutional balancing.
6. Judicial TDM Harmonization: Reflects emerging Text and Data Mining (TDM) approaches adopted in jurisdictions such as the EU, Japan and Singapore, though through judicial interpretation rather than legislation. Enhances India’s attractiveness for responsible AI innovation. Example: Global AI norms.

Limitations & Emerging Concerns

1. Creators receive no assured remuneration despite commercial use of their works.
2. Fair dealing boundaries remain uncertain until final adjudication.
3. Does not address pirated datasets or copyrighted material behind paywalls.
4. Market substitution by AI-generated content may still affect publishers.
5. Policy vacuum persists regarding commercial AI training. Example: News publishers.

Way Forward

1. Enact Explicit Text and Data Mining (TDM) Exception: Introduce statutory provisions with opt-out rights, following the EU model. Example: Copyright amendment.
2. Develop Statutory Collective Licensing: Create a “One-Nation, One-Licence” mechanism (DPIIT Working Paper, 2025) for commercial AI training while ensuring affordable access. Example: Collective Rights Management.
3. Mandate Algorithmic Accountability: Require watermarking, provenance tracking, transparency reports and safeguards against memorisation. Example: AI watermarking.
4. Strengthen Institutional Governance: Align Copyright Act with the IndiaAI Mission, Digital India framework and future Digital India legislation through multi-stakeholder consultations. Example: AI governance.

Conclusion

Technology must empower society responsibly, ANI v. OpenAI demonstrates that evolving copyright through balanced judicial interpretation can simultaneously protect creators while enabling India’s AI leadership.