AI Makes It Easy. Misusing Pitches Doesn't

As AI-generated visuals become routine in creative pitches, a new legal grey zone is emerging, one that many businesses underestimate until it becomes a dispute. A common scenario illustrates this risk. Multiple creators submit pitch materials for a campaign. One creator includes highly tailored AI-generated images as part of their deck. Another creator presents a stronger concept overall and is awarded the mandate. Believing it to be a practical step, the company then forwards the AI-generated visuals from the rejected pitch to the selected creator and asks them to incorporate those visuals into the final output. At first glance, this may appear harmless, especially given the perception that AI outputs are free of copyright protection. But under Indian law, this act can expose the company and sometimes the winning creator to significant legal liability. Let's break this down.

The first layer of complexity lies in the copyrightability of AI-generated works themselves. Indian copyright law is grounded in human authorship. The Copyright Act, 1957 (Copyright Act), defines the "author" of works across categories like literary, dramatic, artistic, musical, photographic, cinematograph films and sound recordings by reference to identifiable humans or legal persons. It does not recognise a machine as an author, nor does it create a special category for "computer-generated works" as some other jurisdictions do. Consequently, where an AI system operates autonomously, with minimal human input or creative direction, the output is unlikely to attract copyright protection in India. The absence of human creativity. 'a core requirement for originality' means that the output falls outside the Act's scope.

However, that does not make AI-generated visuals free for all to use. Copyright protection can arise in the human contribution surrounding the AI output. If a creator meaningfully structures, curates, modifies, selects or integrates AI-generated images into a coherent artistic presentation or narrative, that human contribution qualifies as original authorship. While the individual AI images may not be protected, the composite deck, its layout, its organisation and its textual interplay with AI visuals are protectable. In practice, this means that forwarding even non-copyrightable AI images as part of a curated deck can still infringe copyright in the overall creative compilation. The risk is therefore not eliminated by the uncertain copyright status of pure AI outputs.

Moral rights add a second layer of potential exposure. Section 57 of the Copyright Act grants authors the right to claim authorship and the right to prevent distortion, mutilation or modification of their work in a manner prejudicial to their honour or reputation. These rights apply only to works with human authorship. If a creator's role was limited to entering prompts into an AI tool and accepting outputs without significant creative input, moral rights may not arise. But where the creator curated or integrated AI images into a larger artistic presentation, those components constitute protectable works. If the company subsequently alters, reuses or repurposes that curated presentation without consent, the creator may claim not only copyright infringement but also violation of moral rights. While this claim is fact-specific, companies should not assume that absence of copyright in AI-generated images immunises them from broader legal risks.

The more substantial legal exposure, however, arises from how Indian courts treat pitch materials. Courts have repeatedly held that creative concepts, treatments, structured pitch decks and format notes are shared for a limited purpose, evaluation and cannot be freely reused. In Zee Telefilms Ltd. v. Sundial Communications Pvt. Ltd., the Bombay High Court considered a concept for a television serial pitched to Zee. Although abstract ideas are not protected, the Court observed that the pitch materials contained detailed character sketches, outlines and developed narrative structures. These constituted copyrightable works, and Zee's subsequent commissioning of a similar show without engaging the originators amounted to impermissible appropriation. The Court treated pitch material as protected expression disclosed only for evaluation, not for repurposing.

The Delhi High Court took a similar approach in Anil Gupta v. Kunal Dasgupta, involving a pitched television show concept. The Court held that even where copyright does not strictly protect an idea, pitch materials disclosed for a specific business purpose can carry an implied obligation of confidentiality. Using or sharing such materials beyond that limited purpose can constitute a breach of confidence, entitling the originator to relief. Later decisions across High Courts have consistently reinforced this principle: pitch submissions are not fair game for reuse or adaptation, particularly where the materials contain specific treatments, visual concepts, curated layouts or developed formats.

This principle of confidentiality applies squarely to the scenario of forwarding AI-assisted visuals from a rejected pitch to the selected creator. The fact that the images were produced using AI does not alter the character of the pitch relationship. The company received the materials for evaluation, not for redistribution. Even if the images are non-copyrightable, their inclusion in a curated deck and their presentation within an organised pitch structure often will be. And even where copyright does not apply, confidentiality obligations subsist. An implied obligation prevents the company from sharing or exploiting these materials, unless consent has been obtained from the originator.

The Deshmukh & Co. decisions further underscore an important legal principle: possession of a work does not confer the right to exploit it. The Bombay High Court held that exploitation rights flow strictly from the scope of contractual consent. Applied to pitch scenarios, this means that merely receiving a pitch deck does not give the company the right to reuse or forward its contents. Whether the materials are human-created, AI-assisted or a hybrid, their reuse requires explicit permission. A company that forwards them without consent risks breach of contract, breach of confidence, copyright infringement (where applicable) and unjust enrichment.

Liability also extends to the appointed creator who ultimately uses the forwarded materials. If the selected creator reasonably believes that the company is authorised to provide and authorise use of the visuals, their exposure may be limited. However, if they know or should reasonably realise that the materials originated from a rejected pitch and are being repurposed without consent, they may be implicated. Where the materials contain protectable elements, the selected creator may be liable for copyright infringement. Even where they do not, the selected creator may be considered a participant in a breach of confidentiality or liable in restitution for unjust enrichment derived from confidential pitch inputs.

The broader lesson for organisations is simple but critical: AI does not dilute confidentiality obligations, originality or contractual limitations. If anything, AI heightens the need for clear protocols because its outputs blur traditional copyright boundaries. Companies should adopt explicit pitch policies that set out what use may be made of submitted materials, clarify whether non-selected materials can be reused, and require express consent before any pitch material, AI-generated or otherwise, is shared with third parties. Without such clarity, businesses may find themselves inadvertently stepping into legal disputes over what seems like an innocuous internal workflow decision.

In an era where AI-generated visuals are ubiquitous and often indistinguishable from traditional creative assets, companies must remember that pitch materials, regardless of how they are produced, remain controlled disclosures. They are shared for evaluation, not exploitation. Treating them as free loating resources may save time in the short term, but it exposes organisations to substantial legal consequences. Respecting pitch boundaries particularly in an AI-driven creative ecosystem is not just prudent; it is legally necessary.

Disclaimer: This article is meant for informational purpose only and does not purport to be advice or opinion, legal or otherwise, whatsoever. Pioneer Legal does not intend to advertise its services through this article.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.