Consumers have always judged products by how they feel – the quiet pleasure of a well-proportioned object and the ease of use that good design makes almost invisible. Over the past decades, as life migrated onto screens, this concern for aesthetics expanded into a distinctly digital space: icons, layouts, and animations.

In this environment, design innovation lives in pixels. These designs are more than functional tools. They are the ‘face’ of the brand. They are the primary interface between a company and its customer. Yet, for a long time, it was unclear if these digital assets ‘under the glass’ could be legally protected.

In a new study (Amoncio et al. 2025a), we ask how legal clarification changes firms’ behaviour. Does it lead them to seek more protection for their digital designs? Does it alter how they appropriate these assets – for instance, by selling design patents rather than licensing them? And are these effects stronger in parts of the design space where many look-alike designs coexist?

Our findings provide guidance on how to make innovation policy fit for the digital age and they speak to broader debates on regulating digital markets and new digital assets (Peukert et al. 2020, Presidente and Frey 2022, Crémer et al. 2024, Lutes et al. 2025).

A billion-dollar verdict that changed the rules

The empirical turning point in our analysis is the 2012 Apple v. Samsung verdict in the US. A California jury awarded Apple US$1.05 billion in damages for, among other things, infringement of the digital design of the iPhone home screen.

Figure 1 Apple v. Samsung case

Notes: The left image depicts Apple’s US Design Patent D604305, covering the digital design of its iPhone home screen. The right image shows Samsung’s Galaxy S smartphone interface. At issue in Apple v. Samsung was whether the overall look and feel of the core operating system constituted protectable design and whether Samsung’s implementation infringed Apple’s design patent.

For physical product designs like chairs, bottles, car headlights, the US courts had long recognised design patents as valid and enforceable. For purely digital designs, however, the situation was murkier. Practitioners we interviewed described early digital designs such as graphical user interfaces filings as a ‘leap of faith’ – firms could file for digital design patents, but no court had clearly confirmed whether such designs were actually protectable and enforceable.

The 2012 jury verdict changed that. It sent a strong signal that digital designs could be treated as protectable design patents and enforced with significant economic consequences. Later stages of the litigation refined how damages should be calculated but left the core message intact: digital designs are real assets that can be legally protected and enforced.

How legal certainty reshapes design markets

Intellectual property (IP) protection holds the potential to create value in two ways: (1) by securing the returns from firms’ upfront investment in design creation, shielding them from unauthorised imitation; and (2) by generating income from the appropriation of rights granted to firms. Once granted, rights can be either kept exclusive, licensed, or sold on markets. Legal uncertainty makes both decisions harder.

When it is unclear whether a digital design can really be protected and enforced, it is difficult to know whether the cost of filing is worthwhile, and it is even harder to value the asset in a transaction. Potential buyers or licensees must spend resources on legal due diligence, and they still face the risk that a court might later decide the design is not protectable after all.

The Apple v. Samsung verdict reduced these “is this even a real right?” doubts. Once courts treated digital designs as enforceable with meaningful damages, the due diligence costs of trading them fell. More digital designs should then cross the threshold where it becomes attractive to seek protection and to use patents in transactions.

At the same time, monitoring remains difficult. Even within digital design, some spaces – think emoji icons or reaction buttons – are crowded with near-twin digital designs, while others, such as specialised medical-device dashboards, are relatively sparse. In crowded spaces, small tweaks can blur the line between compliance and infringement, making it costly to police ongoing licences. Selling the asset by rights transfers, by contrast, require due diligence only once and avoid future monitoring disputes because the original creator offloads the burden of policing small, iterative infringements to the new owner.

This yields three testable predictions. First, clearer rights should lead to more protection of digital designs. Second, appropriation should shift toward outright transfers rather than licences. Third, this shift should be strongest in visually crowded parts of the design space where monitoring is hardest.

How firms use and exploit digital design protection

To test predictions, we combine several rich data sources. We focus on design protection in the US during the period 2009–2015, drawing on the universe of design patents examined by the US Patent and Trademark Office. Our unit of analysis is a design subclass in a given year, which allows us to observe how protection and commercialisation evolve within narrowly defined design spaces over time.

A central challenge is distinguishing digital designs, such as graphical user interfaces, icons, and screen-based elements, from more traditional, physical designs. To do so, we follow the USPTO’s own guidance on digital design protection and combine classification information with targeted searches in patent titles and claims. This approach allows us to identify subclasses where digital designs are meaningfully present, while excluding filings that merely depict generic screens without interactive or visual functionality.

To isolate the effects of digital design protection, we then match each digital design subclass to a closely related physical design subclass. The matching is based on similarities in citation patterns, ensuring that digital and physical subclasses are comparable in their technological and market context. The resulting dataset consists of a balanced panel of subclass–year observations covering both digital and physical design domains.

Figure 2 Annual count of digital and physical design protection

Notes: The solid line shows total design applications in digital subclasses. The dashed line represents applications in matched physical subclasses. The red vertical line denotes the 2012 Apple v. Samsung jury verdict. The sample is based on a balanced panel of 21,553 subclass–year observations (2009–2015).

We link these design patents to USPTO assignment records to observe how firms exploit their protected designs. These records allow us to distinguish between outright transfers of ownership and recorded licensing arrangements, providing a window into how design rights are traded and shared across firms. In addition, we trace whether protected designs are commercialised in consumer markets by linking patents to product listings on major online marketplaces. Finally, we use computer vision techniques à la Amoncio et al. (2025b) to analyse the visual similarity of designs within each subclass. This allows us to measure how crowded a design space is, that is, how many close visual look-alikes exist, which captures the difficulty firms face in monitoring and enforcing their design rights.

What we find

First, legal certainty increased digital design protection relative to matched physical designs. After 2012, digital design subclasses see roughly a 9% increase in protection activities compared to their physical counterparts, controlling for common trends. The effect is stronger for designs that are employed in a commercialised product, suggesting that firms responded by investing in protecting economically meaningful digital designs rather than simply clearing out a backlog of old filings.

Figure 3 Event study: Effect of legal certainty on digital design protection

Notes: Each dot shows, for a given year, how much more often firms sought protection for digital designs than for similar physical designs, relative to 2012 (marked by the red dashed line). Values above zero mean that digital designs were being protected more frequently than we would have expected based on pre-2012 trends. The thin vertical lines indicate the range of statistical uncertainty around each estimate.

Second, firms’ appropriation shifted toward transfers, not licences. We observe a statistically and economically significant rise in recorded transfers of digital design patents after the verdict, but essentially no change in licensing activity, which remains rare and highly concentrated. Interviews help explain this pattern. Practitioners described digital designs as being tightly bound up with brand identity. Letting someone else use that ‘face’ is risky: if the licensee uses the design badly or in the wrong context, it can backfire on the original brand, and the owner must keep checking that the design is not being misused. Transfers, by contrast, allow a clean change of ownership with one-time due diligence and no ongoing oversight.

Third, the shift toward transfers is strongest in crowded design spaces. Results show that post-verdict increases in transfers are concentrated in subclasses where many look-alike designs exist. In these crowded spaces, it is harder to spot and prove infringement, so firms are more inclined to sell designs outright rather than license them.

Why this matters for policy

Many US firms have become leaders in the expansion of global tech markets. Digital design increasingly confers a competitive edge, and innovation in ecosystem design and sophisticated user interfaces now constitutes a central dimension of competition. This goes well beyond the functional deployment of digital technologies or the development of new services. Design has become a strategic asset in its own right.

How, then, can policy best support this emerging dimension of innovation? Our findings point to the importance of favourable framework conditions. For firms to become ‘aesthetic pioneers’ and sustained investors in digital design, legal rules must be clear and predictable. We show that legal certainty regarding the scope and enforceability of rights can reshape innovation incentives and encourage design creativity – but along margins that differ from standard expectations.

In particular, clarifying the scope of design protection does not automatically promote diffusion. In markets for processes or components, clearer rights often go hand in hand with more licensing. For digital designs, however, this mechanism breaks down. Appropriation shifts toward outright transfers and ownership changes, while licensing remains rare. In policy contexts where shared design standards or interoperable digital designs are desirable, this implies that broader diffusion and reuse are unlikely to emerge spontaneously through markets alone.

References

Amoncio, E, A Cuntz and C Fink (2025a), “The Future is Under the Glass: Digital Design Protection and Appropriation Strategy”, WIPO Economic Research Working Paper.

Amoncio, E, T Chan and C Storz (2025b), “Using computer vision to measure design similarity: An application to design rights”, Research Policy 54(9): 105309.

Crémer, J, P Heidhues, M Schnitzer and F Scott Morton (2024), “Apple’s exclusionary app store scheme: An existential moment for the Digital Markets Act”, VoxEU.org, 6 March.

Lutes, B, A Cuntz and M Sahli (2025), “Elvis’ ghost, Jordan’s Airness, and the new economics of publicity rights”, VoxEU.org, 16 November.

Peukert, C, S Bechtold, T Kretschmer and M Batikas (2020), “Regulatory export and spillovers: How GDPR affects global markets for data”, VoxEU.org, 30 September.

Presidente, G and C B Frey (2022), “The GDPR effect: How data privacy regulation shaped firm performance globally”, VoxEU.org, 10 March.