When AI Makes The Campaign: Who Controls Fashion’s New Creative Supply Chain?

AI is changing how fashion campaigns are created, adapted and distributed. The next IP challenge may not be the advertisement itself, but everything behind it.

For decades, the fashion campaign had a beginning, a production process and an end. There was a brief. A creative director. A photographer. A model. A location. A production team. A final image. Once the campaign was approved, everyone broadly knew what they were looking at and, importantly, who had created what. Artificial intelligence is beginning to change that certainty.

Generative AI can now assist with campaign concepts, generate images and videos, alter backgrounds and styling, create advertising copy, and produce market-specific versions of a campaign at remarkable speed. For fashion brands operating across multiple markets, the commercial attraction is obvious: more content, faster production and greater localization.

But there is a more important change taking place underneath the efficiency story. The traditional fashion campaign was an asset. The AI-enabled campaign is increasingly a system. And once the campaign becomes a system, the intellectual property questions become much larger than the question of who owns the final image.

From One Campaign to a Creative Supply Chain

Imagine a global luxury brand preparing a new campaign. The original photographs were created in Paris. The brand approves the creative direction. Then the AI workflow begins. The campaign is adapted for India, Japan and the Middle East. The model may change. The background may change. The language may change. The styling may change. The format may change for different digital platforms. The product remains the same.

But is the campaign still the same campaign?

This question matters because AI is not simply producing another photograph. It can participate in several stages of the creative process from creation and editing to localization and distribution. Fashion companies are already demonstrating commercial potential. Zalando reported in 2026 that around 90% of its on-site marketing content was being generated using AI, with content creation time reduced from weeks to days across its markets.

The implication is significant. When one creative idea can produce hundreds of variations, the campaign is no longer necessarily a fixed piece of creative work. It becomes a creative supply chain with inputs, human decisions, technology, outputs and multiple points of approval.

That is where the legal challenge begins.

The IP Question Starts Before the AI Creates Anything

Consider the archive of a luxury fashion house. It may contain decades of campaign photographs, runway imagery, sketches, artwork, celebrity photographs, product designs and other creative material. Some may be owned by the brand. Others may have been commissioned from photographers, designers, agencies or production companies under specific contractual arrangements.

Now suppose the brand uses some of that material to help generate a new AI campaign. The new image may look entirely different from the original. But the first question should not necessarily be “who owns the new image?” It should be whether the brand has the right to use everything that went into creating it. This distinction is becoming increasingly important.

A license negotiated for conventional advertising may not have contemplated every possible form of AI processing. A model release may permit photography and digital advertising without expressly addressing synthetic alteration. An agency agreement may deal with ownership of final campaign materials without explaining what happens when AI creates hundreds of derivative variations. The risk, therefore, may begin before the prompt is written.

The Copyright Act, 1957, provides a starting point. Section 2(d) (vi) identifies, in relation to certain computer-generated literary, dramatic, musical or artistic works, the person who causes the work to be created as the author. But that provision predates today's sophisticated generative-AI workflows and does not by itself resolve every question concerning AI-generated output and human contribution.

The Department for Promotion of Industry and Internal Trade's working paper on the Generative AI-Copyright interface similarly identifies unresolved questions around copyrighted material used as AI inputs, AI-generated outputs, authorship and liability.

The important point for businesses is, therefore, simple; the copyright question exists on both sides of the AI system about what goes in, and what comes out.

Ownership Is Not the Same as Control

The debate around AI often starts and ends with ownership. But for a fashion brand, ownership may be only one part of the commercial equation. Four questions need to be separated:

1. Who owns the asset?
2. Who controls its use?
3. Who has permission to commercially exploit it?
4. Who can stop someone else from using it?

These answers may not always be the same. A brand may obtain contractual rights to use an AI-generated asset while the copyright position in particular AI-generated elements remain uncertain. Equally, a brand may own its own trademarks and visual assets but cannot assume that every AI-generated image inspired by a particular visual aesthetic is automatically safe to use. This becomes especially interesting when AI reproduces the look and feel of a brand without reproducing its logo.

The Trademarks Act, 1999, recognizes advertising as a context in which trademark infringement can arise, while Section 29 (8) specifically addresses advertising that takes unfair advantage of, or is detrimental to, the distinctive character or reputation of a registered trademark.

The future question may therefore move beyond whether the trademark was copied to “Has the AI-generated creative appropriated enough of another brand's commercial identity or reputation to create a legal problem?”

The answer will depend on the facts and the applicable legal doctrine. But in fashion where color, packaging, imagery, silhouettes, campaign aesthetics and other visual signals can contribute heavily to brand recognition, that question deserves serious attention.

The Human Face Creates Another Layer

Fashion advertising has another unique complication: people are part of the creative asset. AI can create a completely synthetic model. It can also modify a real model's face, body, clothing or surroundings, or create an individual who appears remarkably similar to a real person. The issue then extends beyond copyright.

Depending on the circumstances, the use of an identifiable person's image or persona may involve personality or publicity interests, privacy, contractual permissions and other applicable rights. This is particularly important when older models or celebrity agreements are being used for new AI workflows.

A contract negotiated for photography and conventional advertising may not clearly address synthetic replicas or extensive AI manipulation. Whether the existing permission is broad enough will depend on the actual language of the agreement, the use being made and the applicable law.

There is also a commercial dimension. A 2026 study examining AI and human models in fashion advertising found that consumers could show greater skepticism toward AI models, with potential consequences for brand advocacy and purchase intentions. For luxury brands, that matters. Luxury does not sell only one product. It sells craftsmanship, aspiration, identity and authenticity.

A campaign can therefore be efficient from a production perspective and still create a problem for the very sense of authenticity that gives the brand its value.

Localization is where the Risk Multiplies

AI's ability to localize advertising may ultimately be one of its greatest advantages for global fashion businesses. But localization is more than translation. A campaign adapted for India, Japan or the Middle East may involve changes to the model, styling, setting, language, cultural references and presentation of the product.

Imagine that a global brand approves one campaign and AI produces 50 local versions.

1. Who approved those 50 versions?
2. Did the original model's consent cover all of them?
3. Who checked the cultural references?
4. Who checked the third-party rights?
5. Who confirmed that the product was being represented accurately?
6. And if one version creates a legal problem, who is responsible for the global brand, local agency, creative team or technology provider?

At this point, creative production becomes risk allocation. India's 2026 amendments to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, have introduced a more specific framework around “synthetically generated information”, including obligations concerning identification and handling of qualifying synthetic content by intermediaries.

This does not mean every AI-assisted fashion advertisement is automatically subject to an identical labelling obligation. Rather, it reflects a broader regulatory direction: realistic synthetic content is increasingly being treated as an issue of transparency and accountability.

The advertising ecosystem is moving in the same direction. Google announced in July 2026 that advertisers can add AI-related labels to image and video advertising creatives, reflecting emerging disclosure requirements in jurisdictions including India.

For brands, the message is becoming difficult to ignore: AI-generated advertising is moving from creative experimentation toward governed business activity.

An Original Advertisement Can Still Mislead

There is another distinction that deserves attention. An AI-generated advertisement can be completely new and still be misleading. Imagine AI creates a highly realistic image showing a garment with a particular texture, finish or appearance that the actual product does not have.

The copyright question may be: Who created the image?

The advertising question is: What does the consumer believe about the product after seeing it?

The ASCI Code requires advertising to be truthful and claims to be capable of substantiation, and addresses misleading impressions created through visual presentation, implication, omission or ambiguity. India's consumer-protection framework also includes the Consumer Protection Act, 2019, and the 2022 Guidelines for Prevention of Misleading Advertisements and Endorsements for Misleading Advertisements.

AI does not create an exception to these principles. A synthetic image is still an advertisement, and the impression it creates still matters.

The Contract May Become as Important as the Creative

The traditional fashion advertising chain might have looked like:

Brand → Agency → Photographer → Model → Production → Campaign

The AI-enabled chain can look very different:

Brand → Agency → Creative Team → AI Platform → Local Team → Media Platform → Consumer

Every additional layer creates another potential question about rights and responsibility. The advertising agreement may therefore need to travel further than it traditionally has. It may need to address what AI tools can be used, what material can be uploaded, whether information can be retained or used for training, who receives rights in the output, who clears third-party rights, who approves localized versions, how model and likeness rights are handled, and who bears the risk if the final content creates a claim.

In other words, the advertising contract is becoming an AI governance document. The objective should not be to slow AI down. It should be to make sure the brand's legal architecture moves at the same speed as its creative architecture.

The Next IP Audit May Begin with the Prompt

Fashion has always converted creativity into commercial value. AI does not change that relationship. It changes how quickly creativity can be generated, how many times it can be modified, and how widely it can be distributed. That means the traditional question, viz, “who owns this campaign” is becoming too narrow.

The more useful questions may be:

1. What went into creating it?
2. Who had the right to use those inputs?
3. What human decisions shaped the output?
4. Whose identity or brand does it represent?
5. Who approved the localised versions?
6. Who is responsible if something goes wrong?

The future fashion campaign may therefore no longer be one photograph, one film or one advertisement. It may be a continuously evolving creative system. And when creativity becomes a system, IP strategy cannot remain focused only on the final output. The future fashion lawyer may not simply be asked to clear the campaign. They may be asked to help design the rules by which the campaign is created. That is where the next chapter of fashion IP may begin.

References:

1. Department for Promotion of Industry and Internal Trade (DPIIT), Ministry of Commerce & Industry, Government of India, Working Paper on Generative AI and Copyright - Part 1: Balancing AI Innovation and Copyright (December 2025). DPIIT Working Paper on Generative AI and Copyright.
2. Ministry of Electronics and Information Technology (MeitY), Government of India, Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, as amended in 2026 - framework concerning synthetically generated information (SGI). MeitY – IT Rules, 2021 (updated 2026).
3. MeitY, Government of India, FAQs - Amendments to IT Rules, 2021 relating to Synthetically Generated Information (SGI) (10 February 2026) - definition, scope and exclusions relating to SGI. MeitY – SGI FAQs, 2026.
4. Advertising Standards Council of India (ASCI), The ASCI Code for Self-Regulation of Advertising Content in India - requirements concerning truthful, honest and non-misleading advertising. ASCI Code.
5. Central Consumer Protection Authority (CCPA), Department of Consumer Affairs, Government of India, Guidelines for Prevention of Misleading Advertisements and Endorsements for Misleading Advertisements, 2022. CCPA Guidelines, 2022.
6. Google Ads, Updates to AI Labelling Requirements (9 July 2026) - developments concerning labels for AI-generated or AI-modified image and video advertising creatives, including emerging requirements in India. Google Ads – AI Labelling Requirements.
7. Zalando, How Zalando Tells Better Stories: The AI Content Powerhouse (6 May 2026) - reporting that approximately 90% of Zalando's on-site marketing content was being generated using AI and describing the company's move toward faster, more localised content production. Zalando – How Zalando Tells Better Stories.
8. Zalando, Zalando Delivers Strong 2025 Results, Expects Further Acceleration in 2026 Through Scaling AI Innovations - reporting that AI-generated product content increased to approximately 90%, with campaign creation time reduced from six weeks to days and content output increasing by 70%. Zalando – 2025 Results and AI Developments.

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.