India’s legal position on GUI registrability under the Designs Act 

The registrability of GUIs in India has long been a grey area. Sauma Tripathi says that for the law to remain relevant, it must evolve to recognize the nuanced spectrum of human-machine collaboration and to protect innovation – even when it emerges from a prompt rather than a pencil. 

A graphical user interface (GUI) is the visual layout through which users interact with digital systems like software, websites or applications by using elements such as icons, buttons, windows and navigation menus.  

The registrability of GUIs in India has long been a grey area under the Designs Act, 2000. Early on, the designs office granted a handful of registrations, most prominently to Microsoft, for “screen displays and icons” under Class 14-04. However, this was a period of procedural ambiguity as India was still in the process of harmonizing its classification with the Locarno system. Practitioners have since observed that these approvals likely reflected administrative leniency rather than settled policy as the office had not reconciled the statutory requirement of a design being applied to a tangible “article” with “screen displays” under Class 14-04, and upon Indian law conforming to the current edition of the Locarno Classification, a new Class 32 for “Graphic symbols and logos, surface patterns, ornamentation” became available. 

This early permissiveness narrowed sharply after the Controller’s decision in 2014 rejecting Amazon’s application for a GUI design on the grounds that a screen display, visible only when a device is operational, fails to fulfil the requirement of consistent eye appeal. The Controller further reasoned that a GUI lacks independent physical existence, cannot be sold separately and does not embody features of shape or configuration judged solely by the eye. Thus, it does not qualify as an “article” under Section 2(a) or as a “design” under Section 2(d) of the Designs Act.  

Almost a decade later, in March 2023, the Calcutta High Court in UST Global (Singapore) Pte Ltd v. Controller of Patents & Designs (AID No. 2 of 2019) reignited the debate. Setting aside the Controller’s refusal of a design titled “Touch Screen”, the court held that a GUI, though a two-dimensional design visible only when the device is operational, satisfies the definition under Section 2(d) of the Designs Act as it is applied to an article through an industrial process and judged solely by the eye. It further clarified that the visibility of a GUI only during active use does not disqualify it and embedding the design via source code into a microprocessor constitutes a sufficient industrial application. The matter was remanded for reconsideration, briefly raising hopes that GUIs might finally be registrable in India. However, following review, the designs office reaffirmed its earlier stance and, to date, no new GUI registrations have been granted. 

Such persistent uncertainty emphasizes the pressing need for legislative clarification. While the Indian position remains fluid, several international jurisdictions are advancing more rapidly. China has emerged as a forerunner having permitted standalone GUI design registrations since 2014 and, as of June 2021, allows partial GUI protection without hardware linkage which reflects a more modern understanding of software-centric design. The EU has recently updated its design protection framework, replacing the Community Design system with the EU Design system. A key aspect of this reform is the broadening of the definition of “design” to explicitly include digital and virtual designs, such as GUIs.  

AI-generated GUIs: Emerging challenges and the ownership question 

These doctrinal shifts are now being challenged by a more radical transformation, i.e., the rise of AI-generated GUI layouts. In the era of prompt-driven creativity, even the digital interfaces we use to interact with software are no longer hand-designed pixel by pixel. Generative artificial intelligence can now draft entire GUIs producing layouts that mimic human decisions in placement, symmetry, visual hierarchy and accessibility. What once required a human eye is now executed at scale through model training and neural inference. 

This evolution in how interfaces are built opens new conversations within intellectual property law. If a GUI is generated by AI, does it enjoy legal protection? Can it be copyrighted? Can it be registered under the Designs Act, 2000? More pressingly, who owns it? 

The Copyright Act, 1957 offers a starting point. Section 2(d)(vi) defines the author of a computer-generated work as the person “who causes the work to be created”. However, this provision assumes that the human role is active and substantial. If the interface design is entirely generated by an AI tool with the human merely selecting among outputs, the originality and creative control necessary for copyright protection may not be met. 

This aligns with the U.S. case Thaler v. Perlmutter (2025), where Stephen Thaler, a computer scientist, developed a generative AI system named the Creativity Machine and used it to create an image titled A Recent Entrance to Paradise. The U.S. Copyright Office rejected the application because the AI was listed as the author (and Thaler as the owner) and the Court of Appeals for the D.C. Circuit upheld that decision. The court reaffirmed that copyright protection extends only to works of human authorship, rejecting the notion that a non-human entity can qualify as an author. It emphasized that copyright law exists to reward human creativity and intellectual labour and that purely autonomous, machine-generated works fall outside its scope. Though this position aligns with U.S. Copyright Office’s Copyright and Artificial Intelligence, Part 2: Copyrightability report released in February 2025, the debate is far from settled. Also, in January 2025, the U.S. Copyright Office took a more nuanced view in a case involving an image produced via the platform Invoke AI. In that instance, registration was granted for an artwork titled A Single Piece of American Cheese, recognizing the applicant’s creative contribution in curating, prompting and refining the output. This marked departure from the Thaler position suggests that where human direction meaningfully shapes the expressive result, even if rendered by AI, the threshold of authorship may still be satisfied. The decision emphasizes a growing interpretative divide in that AI may not independently own rights yet works generated through guided collaboration between human and machine could fall within the ambit of protectable expression. 

Across the Atlantic, however, the European Union (EU) has approached the issue differently. The recent EU Artificial Intelligence Act does not determine authorship or copyright ownership of AI-generated content and is, at its core, a product-safety and risk-regulation instrument. The act, however, requires providers of general-purpose AI models to adopt a copyright compliance policy aligned with Union law and publish a sufficiently detailed summary of the training data used. While these measures enhance transparency and reduce infringement risk, they stop short of granting or defining authorship or protectability of AI outputs. In essence, the EU framework seeks to ensure lawful data use and accountability but leaves open the question of whether AI-generated GUIs attract IP protection to existing copyright and design regimes. 

Separately, Indian copyright law offers an alternative layer of protection for GUIs. Under the Copyright Act, 1957, copyright subsists in original literary works, which expressly include “computer programmes”. Thus, software source code can be registered as a literary work and such protection would extend to screen displays and GUI elements generated by the program to the extent they are expressed through the code. While this does not protect a GUI as a standalone visual design, it provides meaningful protection against unauthorised copying of interface layouts implemented through software. 

Beyond ownership: The emerging infringement landscape 

Generative AI models are often trained on datasets that include proprietary or copyrighted design elements. If such a system unintentionally reproduces aspects of an existing interface, a user may unknowingly deploy infringing content. Without transparency in training datasets or audit tools to verify originality, the risk of derivative copying looms large.  

The associated commercial uncertainty is significant. If a startup builds its app interface using AI-generated layouts, can it claim IP rights? Can it prevent others from using similar elements? Can it license its UI design or register it for protection? Current law does not offer easy answers. Courts in India have not yet dealt with such cases and regulatory frameworks remain silent. 

It is pertinent to state here that in addition to copyright and design law, certain GUI elements can also attract protection under trademark or trade dress principles. Distinctive app icons, logos, colour schemes or even a recognizable overall screen layout can, in some cases, serve as source identifiers and be protected under the Trade Marks Act, 1999. India’s trademark regime recognizes non-traditional marks and courts have also extended passing-off protection to get-up or the visual appearance of products where goodwill/reputation and likelihood of confusion are shown. However, this protection is narrow as GUI elements must be non-functional, and sufficiently distinctive or have acquired secondary meaning. 

In practice, clients and creators must increasingly rely on contractual safeguards. Clear and well-drafted terms of use with AI service providers are critical as many platforms either restrict commercial reuse or reserve ownership rights over generated outputs. Designers and developers using such tools should maintain detailed records of prompts, output iterations and post-processing to evidence human contribution. For clients investing in proprietary user interfaces, legal due diligence should encompass an assessment of originality, the traceability of training datasets, clarity of output ownership under platform terms and the feasibility of registering the GUI under applicable IP regimes, etc. 

The road ahead: Bridging innovation and legal ownership 

The commercial weight of GUI design makes the existing legal ambiguity increasingly difficult to justify. According to analysis by Market Research Future, in 2024, the Indian GUI design software market was valued at approximately US$1.78 billion, forming part of a global market valued at about US$25.5 billion. Furthermore, both markets are projected to grow at a similar pace of around 10 percent CAGR over the next decade with India expected to reach approximately US$5 billion by 2035 and the global market expanding to about US$73 billion over the same period. These numbers make it clear that GUIs are not just aesthetic layers, they are fast becoming a core business asset shaping brand identity, user engagement and customer trust. And yet, despite this economic momentum, the absence of clear IP protection leaves a growing gap between innovation on screen and ownership in law. 

But these gaps point to a larger legislative vacuum. At present, India’s IP laws are silent on machine authorship or human-machine collaboration. As AI becomes a standard tool in creative and technical domains, this vacuum will increasingly limit protection and create commercial risk. There is scope, therefore, for introducing statutory presumptions that attribute limited rights to the human operator or prompt-giver where sufficient creative contribution can be demonstrated even when the final output is AI-rendered. Similarly, the Designs Act could be updated to recognize interface designs independent of underlying hardware, following models already adopted in China and discussed within the EUIPO. 

In conclusion, AI-generated GUI layouts represent the bleeding edge of design innovation. Convenient, scalable and functional but legally precarious. Without human involvement, protection is elusive. Without legal reform, ownership remains uncertain. And without careful contracting and due diligence, infringement risk can silently creep in. That said, Indian policymakers are aware of the gaps between existing statutory frameworks and modern, software-driven design and discussions around updating the law to accommodate GUI protection and AI-assisted creativity are already underway. For the law to remain relevant in this new creative paradigm, it must evolve to recognize the nuanced spectrum of human-machine collaboration and to protect innovation, even when it emerges from a prompt rather than a pencil.