Media, creative agencies' AI push under copyright spotlight, faces policy uncertainty

India's media and creative agencies are pouring investments into generative artificial intelligence—from dedicated AI studios and automated production units to synthetic content pipelines—as they race to meet clients' growing demand for faster and cheaper advertising. But legal experts warn that while AI may dramatically reduce production costs, it does not shield agencies from copyright liability, leaving the industry's AI-led business model vulnerable to future policy changes.

Global holding companies including WPP, Publicis Groupe, Omnicom, Dentsu and Havas have launched AI-powered creative platforms globally, while their Indian operations have increasingly embedded AI across copywriting, visual production, media planning, localisation and performance optimisation. Independent agencies too are setting up dedicated AI capabilities to meet growing client demand for faster campaign execution and lower production costs.

Industry executives say AI is helping agencies compress production timelines from weeks to days, automate repetitive creative tasks and produce multiple campaign variants across languages and platforms. As clients seek greater marketing efficiency amid pressure on advertising budgets, agencies increasingly view AI as a competitive necessity rather than an experimental technology.

However, legal experts say the industry's AI ambitions now face an equally significant challenge: India's evolving copyright and AI policy framework.

Unlike software automation, generative AI raises questions about ownership, authorship and liability—issues that could determine whether agencies retain the commercial benefits of their AI investments or face rising compliance costs and litigation.

Who owns AI-generated advertising?

The legal uncertainty begins with authorship.

Under Section 2(d)(vi) of the Copyright Act, 1957, the author of a computer-generated literary, dramatic, musical or artistic work is "the person who causes the work to be created."

That means when an agency employee generates advertising assets using an AI platform, the agency cannot escape responsibility by arguing that the machine created the content.

"As media and creative agencies invest in AI studios to produce advertising content, they may increasingly come to bear legal responsibility for that content," legal experts said.

The Indian Copyright Office has also maintained that copyright registration requires human authorship, meaning AI systems themselves cannot qualify as authors.

Depending on contractual arrangements with advertisers, agencies may also become the first owners of copyright for commissioned commercial works, making them responsible for infringement claims arising from AI-assisted campaigns.

AI doesn't create a copyright loophole

Despite rapid advances in generative AI, experts believe agencies cannot treat AI as a legal shield.

"Agencies cannot simply bypass India's copyright laws by relying on AI," said Bharadwaj Jaishankar, Partner at CMS INDUSLAW.

"Under the Copyright Act, the agency deploying the AI tool remains liable for infringement. However, pure AI outputs, with no human involvement, currently fall into a grey area and are arguably not covered by copyright at all as the scheme of the Act envisages human authorship."

The responsibility begins well before an advertisement reaches consumers.

Agencies must ensure that datasets used to train or fine-tune AI systems—including photographs, artwork, music, videos and written content—have been lawfully licensed.

They must also verify that AI-generated outputs do not substantially reproduce protected third-party works.

These questions are now central to the closely watched Delhi High Court case between ANI Media Pvt. Ltd. and OpenAI, which is examining whether using copyrighted works to train AI models without permission amounts to copyright infringement. The judgment, which has been reserved, is expected to become one of India's most significant legal precedents for generative AI.

Fair dealing unlikely to protect agencies

Legal experts also believe agencies are unlikely to find protection under India's fair dealing provisions.

Section 52 of the Copyright Act permits fair dealing only for purposes such as private use, criticism, review and reporting of current events.

Commercial advertising falls outside these exceptions.

Even contractual agreements with AI vendors offer only limited comfort.

While agencies may negotiate indemnity clauses with AI providers, these arrangements merely shift financial liability between contracting parties and cannot extinguish statutory rights available to copyright owners.

A limited defence may exist where copying is trivial or insignificant, but experts say that threshold is unlikely to protect agencies deploying AI at commercial scale.

Due diligence becomes a business imperative

The legal risks are no longer theoretical.

"Media and creative agencies are equally expected to undertake adequate due diligence before deploying AI-generated content," said Sonam Chandwani, Managing Partner at KS Legal & Associates.

"Existing copyright jurisprudence imposes obligations upon commercial entities to ensure that intellectual property rights are respected before publication. An agency that ignores obvious similarities with pre-existing works or fails to implement reasonable verification processes may find it difficult to escape allegations of negligence, contributory infringement or inducement of infringement."

She added that contractual disclaimers or AI vendor agreements cannot override statutory copyright protections.

"While agencies may seek indemnities from AI technology providers, such contractual arrangements merely redistribute financial risk between contracting parties and do not extinguish the rights of an aggrieved copyright owner to proceed against the agency that commercially exploited the infringing work."

For agencies, this means AI governance is likely to become as important as creative capability. Several multinational networks have already introduced internal AI policies governing prompt engineering, disclosure requirements, approval workflows and human review before AI-generated content is released to clients.

Maninder Adityaraj Singh, Innovation Head at Rediffusion, said the global conversation around AI in advertising has already shifted from experimentation to accountability.

"The direction is clear: AI isn't being stopped, but it is being made accountable," Singh said. Pointing to developments such as the European Union's AI Act and evolving AI governance discussions in India, he said agencies would increasingly be expected to demonstrate how AI-generated work is created, reviewed and commercialised.

According to Singh, compliance is becoming an integral part of the creative process itself. Agencies will need to understand which AI tools they use, whether training datasets or creative references involve third-party intellectual property, whether voice and likeness rights have been authorised, and whether AI-generated outputs are safe for commercial deployment.

"There will be a compliance cost through AI policies, legal review, licensing, documentation and employee training," he said. "But that's the new cost of trust. The real cost will be non-compliance—whether through copyright disputes, personality rights violations or reputational damage."

He added that agencies investing early in responsible AI governance would be better placed to assure clients they can deliver campaigns not just faster, but with greater accountability and brand safety.

Rahul Vengalil, CEO and Co-founder of tgthr, believes those compliance requirements could create a divide within the industry.

"The cost of AI copyright compliance is significant. Large agencies have legal teams, governance processes and audit mechanisms to manage it, but many smaller agencies and startups have historically been lax about copyright compliance, with or without AI. That divide is unlikely to disappear. Some companies will invest because the legal and reputational risks are too high, while others may continue to cut corners until enforcement becomes stricter," Vengalil said.

Will policy changes alter agency economics?

Experts believe India's eventual AI framework will determine whether agencies can continue scaling AI-led production without significant regulatory friction.

"What India needs is balance," said Himanshu Deora, Partner at King Stubb & Kasiva.

"Practically, that could mean greater transparency around what data AI models are trained on, clearer rules on ownership and authorship for AI-assisted work, defined liability standards across AI developers and businesses using their tools, disclosure norms when AI-generated content is used in commercial advertising and practical licensing mechanisms that allow copyrighted material to be used lawfully rather than everyone operating in a grey area."

According to Deora, regulatory certainty would encourage innovation while protecting creators' rights.

Compliance could move upstream

Future regulations may also shift responsibility away from digital platforms and towards advertisers and agencies themselves.

Astha Sharma, Partner at AQUILAW, said policymakers should close what she described as an enforcement gap under the Information Technology Rules.

"The technical burden of labelling synthetically generated content currently falls almost entirely on digital platforms. Future frameworks should instead impose direct obligations on media agencies and brands commissioning advertisements."

She said agencies could eventually be required to disclose AI-generated commercial content directly to the Advertising Standards Council of India.

Sharma also suggested regulators consider creating a safe harbour for agencies that maintain detailed prompt logs and use enterprise AI models trained exclusively on licensed datasets.

Equally important, she said, is clarifying what level of human creativity is necessary before AI-assisted works qualify for copyright protection.

Drawing from the Supreme Court's ruling in Eastern Book Company vs D.B. Modak, Sharma said merely entering a prompt into an AI system may not satisfy the threshold of skill, judgment and creativity required under Indian copyright law, whereas extensive human editing, localisation and creative intervention could.

A defining moment for the agency business

For India's advertising industry, the stakes extend well beyond intellectual property law.

Agencies have spent the past two years positioning AI as the next engine of productivity—reducing production costs, expanding creative output and enabling personalised marketing at scale.

But if future regulations impose mandatory disclosure requirements, licensing obligations, stricter due diligence standards or greater liability for AI-generated campaigns, the economics of AI-powered advertising could change substantially.

Naresh Gupta, co-founder of Bang in the Middle, said the commercial use of AI-generated content remains legally ambiguous because the technology can closely mimic existing creative works.

"AI can generate work that closely resembles an original creation and modify it based on prompts. Who owns that output and whether it can be commercially used without a licence remains a grey area," Gupta said.

He expects legal clarity to emerge only once Indian courts begin adjudicating copyright disputes involving generative AI. "We haven't yet seen a major copyright case involving AI-generated advertising, but I don't think it's far away. Court rulings will eventually provide the clarity the industry needs."

Gupta also cautioned brands against assuming AI-generated imitations are free from legal risk. "With personality rights receiving greater protection, even AI-generated voices, mannerisms or distinctive styles associated with well-known individuals could expose brands to legal claims. Advertisers need to tread carefully."

For now, agencies continue to invest aggressively in AI capabilities. Whether those investments become a lasting competitive advantage—or evolve into a complex compliance challenge—will depend largely on how India's courts and policymakers define the rules governing AI and copyright over the coming months.