Data Licensing: Why Our Existing Models Fall Short

Why is data licensing so difficult when technology licensing has long relied on established models and commercial frameworks? In MedTech and life sciences, data is becoming increasingly central to product development, validation, regulation and AI-driven innovation, yet familiar licensing logic does not translate easily.

This article builds on the discussion in the MedTech Thought Leadership panel at the LESI 2026 Annual Conference, moderated by Jennifer Burdman, with panelists Rebecca Hays, Chitra Kalyanaraman, Ksenia Takhistova and myself. The conversation highlighted how differently data is understood across industries. This article develops that discussion further from a technology licensing perspective, focusing on what happens when data and licensing logic meet.

Why Data Does Not Fit Traditional Licensing Logic

In technology licensing, we are used to strive for clarity. The asset is defined, the rights are clear, and the scope of use can be described with reasonable precision. Once these elements are in place, licensing can scale. That is why we have been able to build global licensing programmes in areas such as connectivity, consumer electronics and other technology-driven sectors.

In MedTech and life sciences, that clarity is more challenging to achieve. Data is becoming a central asset in how products are developed, validated and used. It plays a role in diagnostics, treatment optimisation, regulatory submissions and increasingly in AI-driven innovation. Yet when we try to apply familiar licensing models to data, something does not quite fit. This is not because data would be a new asset class. It is because data behaves differently from the assets we are used to handle in licensing.

Data Behaves Differently as a Licensable Asset

From a technology licensing perspective, a licensable asset is typically something that can be clearly defined, transferred or accessed in a predictable way, and used by the licensee to create value. Data challenges each of these assumptions. It is not static, but evolves over time. Its use is conditional, often limited by consent and regulation. Its value depends on context and on how it is combined with other datasets. Perhaps most importantly, ownership of data does not define what can be licensed. Permission for use defines the scope.

This shift is more significant than it may first appear. In traditional IP licensing, once ownership is established, the focus moves to how rights are granted and monetised. In data licensing, the question of whether the asset can be used at all, and for which purposes, remains central throughout the lifecycle of the agreement. That alone limits scalability and complicates deal structures.

Data Is Not One Asset but a Stack

Another source of confusion is that in many transactions, “data” is treated as if it were a single asset. In practice, it is better understood as a stack. There is

• the device that generates the data

• the dataset itself

• the platform that aggregates

• the algorithm that interprets it, and

• the insights or outputs that are ultimately used

Each of these layers has different characteristics, different value drivers and often different legal treatment. When these layers are not clearly separated in the transaction, misunderstandings are almost inevitable. Many negotiations may stall because the parties are not aligned on what is actually being licensed, not because they would disagree on the terms.

Why Valuation Becomes Difficult

This lack of clarity becomes particularly visible in valuation discussions. Technology-driven organisations tend to look at scale, reuse and integration potential as basis of valuation of the asset. Life sciences organisations focus on quality, reproducibility and regulatory acceptance. These are not simply different metrics. They reflect fundamentally different views on what creates value. One side is asking how broadly the asset can be used, while the other is asking whether it can be relied upon in a regulatory context. Attempting to bridge this gap purely through pricing discussions is rarely effective if the underlying definition of value has not been aligned.

Data Licensing Gathers Complexity

The same challenges are reflected in deal structures. Existing templates, whether software licences, research collaborations or co-development agreements, capture parts of the reality but not the whole. Data licensing in MedTech and life sciences sits somewhere between technology licensing and the transfer of a regulated asset. As a result, agreements tend to become hybrid structures, often complex and highly tailored. This contributes to the fact that most data-related transactions remain bespoke and therefore relatively slow to negotiate.

From a broader perspective, this also explains why data licensing has not yet reached the level of scalability seen in some other sectors. Licensing tends to scale when assets are clearly defined, rights are transferable and value is predictable. In telecommunications, for example, standardisation created a common framework that enabled large-scale licensing programmes. In the context of MedTech data, we are still far from that point. Datasets are fragmented, rights are inconsistent, value promise remains open-ended, and regulatory constraints vary across jurisdictions. There is no common framework that would support standardised approaches.

Data Will Complement, Not Replace, Traditional IP

At the same time, it is important that we do not see this development as a shift away from traditional IP. Data is not replacing patents or technology. Instead, it is becoming a complementary asset. Patents provide freedom to operate, technology enables practical implementation, and data increasingly drives performance, optimisation and validation. Future licensing strategies will need to combine these elements rather than treat them separately.

For licensing professionals, this does not require abandoning familiar principles, but it does require re-examining some core assumptions. The fundamental questions remain the same. What is the asset? What rights are being granted? Can the licensee actually use it? Where does the value come from? The difference is that the answers are less straightforward and more dependent on context.

Building New Frameworks for a Different Kind of Asset

We are, in effect, trying to license something that behaves like a product, is regulated like a drug, and is used like software. It is therefore not surprising that existing models struggle. The opportunity is in adapting current approaches, and in developing new frameworks that reflect the nature of the asset itself.

That is why data licensing should be a shared point of discussion among licensing professionals. Its significance goes beyond individual transactions. Data has the potential to reshape licensing frameworks, challenge established assumptions and drive the development of new approaches across the field.