The question of the shift in value within digital infrastructures

The rise of artificial intelligence is gradually transforming digital infrastructure: it is becoming more distributed and integrated, while increasingly bringing together computing power, networks, data, and physical systems. This evolution is changing the nature of systems and how they are integrated, but it also raises questions about how value is created, distributed, and captured within technological ecosystems.

The OECD describes digital infrastructure as “connective tissue” that enables digital systems to function, interact, and scale. This interconnecting function becomes all the more important as AI spreads throughout increasingly distributed systems, particularly with the development of Edge AI and Physical AI. Moreover, this issue is not limited to AI alone. It also applies to quantum technologies and the digital infrastructure needed to harness them, as well as, in the future, to hybrid infrastructures combining AI, high-performance computing, and quantum computing.

As these technologies are integrated into increasingly complex systems, a question arises: how are the determinants of value creation, capture, and sharing evolving?

A reshaping of the determinants of value

In digital infrastructure, value can still be relatively easily linked to specific building blocks: components, software, platforms, networks, or services. However, the growing integration of AI makes this interpretation less obvious and likely less relevant. Indeed, with the development of Edge AI and Physical AI, the functionalities and capabilities enabled by AI no longer rely solely on centralized models and infrastructure. They are gradually being distributed across physical systems, vehicles, robots, machines, networks, and operational infrastructure.

This evolution also changes the role of data within these systems. Data is no longer merely a resource used by AI models: its collection, location, quality, flow, sharing, processing, and protection have themselves become key factors in the ability to create value.

In distributed infrastructures, particularly at the edge, part of this value may thus be linked to the ability to generate, process, and leverage data as close as possible to its origin.

Thus, the value of an intelligent system depends not only on the model or components that make it up, but also on how these integrate with data, sensors, computing systems, networks, software, actuators, and infrastructure.

A similar logic can be observed in quantum technologies. Value does not lie solely in the QPU, but in the full range of capabilities required to harness it: control, interconnections, electronics, software, algorithms, error correction, classical computing, the cloud, and applications.

In both cases, this evolution leads us to reexamine the determinants of value: value no longer necessarily resides solely on technologies considered in isolation, but also results from their integration and coordination within complex digital infrastructures, as well as from the resources—particularly data—and the capabilities needed to harness them.

It is therefore not necessarily a matter of a transfer of value from one technology or layer to another. Rather, it is a matter of understanding how the conditions under which value is created and captured are evolving.

Data Spaces: an infrastructure that also defines the conditions for value creation

Data spaces illustrate this evolution particularly well. A data space is not merely a technical mechanism for sharing data. It combines data infrastructures and governance frameworks to enable the sharing and reuse of data under secure and interoperable conditions. Viewed as an infrastructure, a data space therefore helps define several essential conditions: who can access which data, under what conditions, for what purposes, and according to which governance rules. But it can also help determine who can create value from this data and how that value can be distributed and captured.

The choice of a data infrastructure thus also becomes a choice regarding the economic terms of its operation. This observation extends beyond data spaces alone. It can be applied more broadly to digital infrastructures that organize access, interoperability, and the integration of multiple resources and capabilities.

When the determinants of value change, governance mechanisms take on strategic importance. Standards do not merely define technical specifications; they help establish the conditions for interoperability among different technologies, different stakeholders, and different layers of an infrastructure.

The experience of the telecommunications sector provides a particularly instructive precedent in this regard. The 3GPP demonstrates how a cooperative organization involving industry players and other stakeholders can help structure a global technology ecosystem around common specifications. The evolution of digital infrastructures can thus give rise to new forms of cooperation and standardization. This issue goes beyond the mere production of standards; it also concerns intellectual property models, conditions for access to technologies and data, and mechanisms for value sharing.

Toward new models of intellectual property and value sharing?

Digital infrastructures already combine very different models: patents, standards-essential patents and FRAND licenses, proprietary software, open source, IP cores, open standards, and hybrid models. It would be simplistic to pit these different models against one another. On the contrary, they can coexist within the same infrastructure, depending on the technologies involved, the nature of the contributions, and the objectives pursued.

The question is therefore no longer just: Who owns the technology?
It also becomes: How should we organize the contribution, access, and sharing of the value created by an infrastructure in which multiple stakeholders participate?
This question does not concern only technologies and intellectual property. It also concerns the conditions for accessing data, its circulation, reuse, and governance. When multiple models of contribution and dissemination must coexist—proprietary technology, FRAND licenses, open source, open standards, data sharing, or hybrid models—technology governance becomes inextricably linked to the question of the economic model and value sharing.

Signs of a shift in technology governance

Several recent initiatives illustrate this shift, at very different levels and based on very different approaches.

The Common European Data Spaces are a prime example. They combine data infrastructures, governance rules, interoperability mechanisms, and conditions for access and reuse. They thus exemplify a form of governance that focuses not only on a specific technology but also on the organization of a data ecosystem.

The World Data Organization (WDO) is another indicator, centered on international data cooperation and governance.

The Artificial Intelligence Infrastructure Interchange (AIII), launched by WIPO, adds another dimension: the evolution of intellectual property infrastructure in the context of AI, particularly regarding data, identification and attribution, rights management, and standards.

The World Artificial Intelligence Cooperation Organization (WAICO), for its part, illustrates the rise of new forms of international cooperation around AI.

Pax Silica brings a different dimension: that of cooperation around strategic value chains for AI, semiconductors, critical raw materials, and supply resilience.

These initiatives are obviously not part of a single framework. They serve different objectives, scopes, and institutional rationales. However, they illustrate a common trend: technology governance is moving beyond the level of individual technologies to focus on the ecosystems in which they are developed, integrated, operated, and leveraged.

The question then becomes one of governing the technology ecosystem itself: how can we organize the interactions between technologies, data, infrastructure, organizations, and capabilities, while enabling the creation, capture, and sharing of value?

From ecosystem to exosystem

This development ties in with another line of thinking we have pursued regarding the concept of the exosystem. The choice of this term is deliberate. It is not simply a matter of considering the technological ecosystem and the relationships among its actors, but of examining the environment that shapes these interactions. In a previous article, this concept led us to conclude that the capabilities necessary for innovation extend beyond the boundaries of any single organization. They involve companies, research labs, industrial partners, standards bodies, open-source communities, academic institutions, and other stakeholders. This perspective can now be expanded: the exosystem also includes the infrastructure, standards, intellectual property frameworks, data infrastructure, access rules, public policies, and cooperation mechanisms that shape the interactions among the components of a technological ecosystem. Technology governance, therefore, is not limited to the organization of the ecosystem. It also encompasses its exosystem—that is, the environment that shapes its functioning and evolutionary trajectories.

The Common European Data Spaces, the WDO, the AIII, the WAICO, and Pax Silica can thus be viewed as very different manifestations of this evolution of the technological exosystem. If value creation increasingly depends on interactions between technologies, data, infrastructure, organizations, and capabilities, then the relevant unit of governance may no longer be the technology or the organization taken in isolation, but rather the ecosystem and exosystem within which these capabilities interact.

The hypothesis of a future AIPP

Based on these indicators, a forward-looking hypothesis can be formulated.

Could the evolution of digital infrastructure gradually give rise to new forms of cooperation and standardization comparable, in some respects, to what the 3GPP has enabled in telecommunications?

One could imagine, for example and looking ahead, a future AI Partnership Project (AIPP) bringing together manufacturers, developers, equipment suppliers, infrastructure operators, academia, and public authorities around common specifications, interoperability mechanisms, and, potentially, new frameworks for intellectual property, data governance, and value sharing.

This would not necessarily involve replicating the 3GPP. Rather, the challenge would be to determine whether the evolution of digital infrastructures—particularly those that will integrate AI, high-performance computing, and quantum technologies—will require a new form of cooperation tailored to their specific characteristics.

Such a scenario raises several questions:

• Which technologies would be likely to be covered by common standards?
• Which stakeholders would participate in defining them?
• What intellectual property mechanisms would accompany these standards?
• What role would SEPs and FRAND licenses play?
• What role would open source and open standards play?
• How should access to, circulation of, and reuse of data be organized?
• How can open contributions, intellectual property, and return on investment be reconciled?
• How should value be shared among the stakeholders who contribute to the infrastructure’s various resources and capabilities?
• And above all: who will define the rules governing how this value is created, captured, and distributed?

Toward a governance framework for digital infrastructure?

This discussion is a direct continuation of the three previous articles in this summer series.

The first led us to shift our focus to the capabilities necessary for innovation and their distribution within an exosystem.
The second applied this framework to computing infrastructures and raised the question of the possible emergence of new families of SEPs.
The third examined the shift in the relevant unit for managing innovation when technologies, resources, standards, business models, and public policies become significantly interdependent.
This new article takes it a step further: it examines how this evolution transforms the determinants of value and, with them, the conditions for its creation, capture, and sharing.

As digital innovation increasingly takes shape within complex digital infrastructures, the governance of these infrastructures itself becomes a key component of innovation capacity.
Rules governing access, interoperability, intellectual property, data sharing, and cooperation directly influence stakeholders’ ability to create, integrate, and evolve new solutions.

Technological governance is thus gradually becoming a matter of governing digital infrastructures and their technological exosystem. It thereby directly helps shape the conditions for collaborative innovation, in a framework that aligns closely with the principles of open innovation.

These issues also pertain to technological sovereignty and resilience. Control over infrastructure, data, critical technologies, and the rules governing their access, interoperability, and evolution directly contributes to the ability of ecosystems to preserve their autonomy, reduce critical dependencies, and withstand disruptions.

The challenge, therefore, is no longer simply a matter of determining which technologies will be strategic.

It is also a matter of understanding which resources and capabilities will be decisive, who will participate in defining the digital infrastructure of the future, according to what rules, and through what mechanisms for creating, capturing, and sharing value.

This may well be one of the strategic questions of the coming years: not only what digital infrastructures we will be able to build and operate, but also what forms of governance we will be able to put in place to organize their development and the conditions for creating and sharing value.