AI supercharges biobanking: Governance and trust must keep pace

Artificial intelligence (AI) is accelerating a structural shift in biobanking, from static repositories of specimens to dynamic, data-rich infrastructures at the nexus of healthcare and research. A new analysis published in Frontiers in Digital Health maps how AI is changing the sector’s operating logic and governance by reshaping four parameters: size, site, access, and speed.

The research “How the world of biobanking is changing with artificial intelligence,” argues that biobanks now function as boundary objects that convert samples and data into assets within a data-driven bioeconomy - an evolution made more urgent by the rise of AI across clinical and research workflows.

How is AI changing the scale and location of biobanks?

The report positions “size” as more than a numbers game. Statistical power still matters, but the utility of data and samples ultimately depends on quality, provenance, and uniform standards - the difference between robust analytics and a classic “garbage-in, garbage-out” problem. In AI contexts, some algorithms can thrive on smaller, high-quality datasets, while deep learning remains data-hungry; either way, quality management is central to value creation.

“Size” also carries strategic and political weight. Large collections can be mobilized to justify funding, signal national capability, and cultivate stakeholder buy-in. As Mayrhofer notes, the category of size is routinely leveraged to validate investments, promote a biobank’s assets, and frame access strategies, evidence that scale can be both analytical and symbolic capital.

The parameter of “site” has expanded beyond the walls of a single facility. Biobanks now operate as sociotechnical assemblages, stand-alone entities, clinical hubs, and federated networks, that stabilize through shared practices, technologies, and norms. In genomics, cross-border data sharing is becoming foundational for precision medicine, exemplified by coordinated ambitions to share large genomic datasets across European countries. In this model, biobanks transform the materiality of samples and data into assets with biovalue, and AI increases the returns to coordinated, transnational infrastructure.

Under the hood, the paper frames the shift with three concepts, assetization, biovalue, and boundary objects, to explain how AI helps convert biological materials and data into economic and scientific value. The analysis underscores that biobanks sit at the crossroads of health research and care, translating multi-stakeholder inputs into shared infrastructure that produces value in a data-driven health economy.

Who can use health data and on what terms?

The access question has moved center stage. Health data practices increasingly feed a “data-driven health economy,” but the same mechanisms that promise precision medicine also raise legitimacy concerns if welfare-state data collection is repurposed primarily for private gain. The paper highlights this governance paradox and calls for access regimes that protect citizens while enabling research and innovation.

In Europe, proposals such as the European Health Data Space (EHDS) reflect efforts to standardize access while balancing rights, innovation, and sovereignty. The review argues that FAIR principles remain vital but must be extended and operationalized to address ethics, incentives, and trust at scale. Without careful design, the drive for more data risks tipping toward an asymmetrical bioeconomy in which citizens are treated as commodified sources rather than empowered participants.

The study also shows how AI intensifies classic access dilemmas. Algorithms amplify the value of provenance, interoperability, and governance guardrails. As biobanks and linked infrastructures grow more complex, issues of trustworthiness, intellectual property, and data sovereignty become more salient. Designing access models that reward sharing while preserving accountability is now a core precondition for responsible AI-enabled biobanking.

Does speed help or harm in AI-enabled biobanking?

The parameter of speed has become a defining pressure point. With national AI strategies and major public-private investments accelerating competition, health systems face strong incentives to adopt AI tools rapidly. The paper notes that speed promises tangible benefits - workflow efficiency, imaging acceleration, and digital biobanking advances in oncology, yet it also widens the gap between fast-moving technology and slower regulation.

Against that backdrop, the study underscores a broad consensus: human-in-the-loop and human-in-command approaches are necessary to maintain accountability and public trust. High-level AI guidelines must be translated into practical processes that biobanks and clinical partners can adopt consistently, even as the technology evolves. In short, speed must coexist with procedural oversight.

Biobanks have always been adaptive, shaped by standards, regulations, and societal priorities, but AI has accelerated datafication and raised the stakes of getting governance right. Where these infrastructures once simply stored specimens, they now co-produce knowledge, rules, and value across clinical, private-sector, and patient-advocacy stakeholders.

What happens next: quality, governance, and equitable value

The study also provides a concrete operational roadmap. To begin with, it puts quality before quantity: invest in data and sample standards, document provenance, and connect networks that can maintain uniformly high levels of quality; otherwise, even large datasets become low-value assets for AI.

Second, build access frameworks for legitimacy. Extend FAIR to real-world policy and practice, clarify rights and responsibilities, and align incentives so that contributors and publics benefit alongside researchers and industry. This requires durable mechanisms for transparency, auditability, and recourse, especially as AI systems scale across borders and sectors.

Third, match speed with oversight. Convert high-level trustworthiness principles into routine procedures, ensure human accountability in sensitive workflows, and sequence deployments where benefits are clearest, such as imaging and oncology, while monitoring for bias, exclusion, and unintended redistribution of value.

Additionally, recognize that AI magnifies biobanks’ role as boundary objects within the bioeconomy. The sector’s future will be determined less by isolated technological wins and more by whether infrastructures can redistribute benefits fairly, maintain public trust, and demonstrate measurable gains for health systems and science. That outcome hinges on getting size, site, access, and speed right, together.