Why this conversation matters now
World Health Expo (WHX) (previously Arab Health) has become a useful barometer for how healthcare innovation is actually unfolding in the region. What is increasingly clear is that the debate has moved on, as discussions are no longer framed around whether technologies and capabilities such as AI and genomics might be adopted, but how they are being deployed, governed and scaled within existing healthcare systems.
In that context, genomics and artificial intelligence are being presented as established components of healthcare strategy in the UAE, rather than emerging concepts. The focus is on execution, which raises practical questions of implementation, particularly around who controls the data, how long it can be used, how outputs are commercialised, and what regulatory constraints apply as projects mature.
Saudi Arabia and other GCC states have announced or implemented genomics initiatives, but the UAE currently stands out as the jurisdiction where population genomics is most clearly embedded into healthcare delivery and underpinned by a deliberately evolving legal and regulatory framework.
Why genomics matters in the Middle East
Genomics has particular relevance in the Middle East because population characteristics materially affect disease prevalence, genetic risk profiles and treatment response in ways not fully reflected in existing global datasets. Founder effects and higher rates of consanguinity increase the incidence of inherited conditions and alter how certain diseases present across populations.
In the UAE, this has translated into a policy decision to integrate genomics into routine healthcare delivery rather than treat it as a discrete research activity. National programmes such as the Emirati Genome Program are structured to link genomic data with clinical records and longitudinal health outcomes. The datasets being generated are intended to support ongoing clinical use, research and innovation over extended time horizons.
This is a critical distinction. Genomic data in the UAE forms part of the national healthcare infrastructure rather than a discrete time-limited project.
Why AI changes the equation
AI is essential to extracting value from population-scale genomics. Machine learning tools support variant interpretation, predictive risk modelling, drug discovery and population health analysis at a scale that would otherwise be unmanageable.
However, AI also changes the legal and governance profile of genomics projects. Model training typically involves iterative and exploratory use of genomic data, often with future applications that are not fully defined at the outset. At the same time, once models are trained, significant value can be generated without further access to the underlying data. In practice, this blurs the boundary between genomic data and the insights derived from it.
It also raises difficult questions around consent, scope, secondary use and control over outputs that may be technically detached from the original dataset but still fundamentally dependent on it. These questions are already material in the UAE, particularly as projects move from pilot to scale.
Legal and regulatory challenges specific to the UAE
One reason the UAE is particularly interesting in this space is that genomics is treated as a regulated healthcare activity and not merely as a category of sensitive personal data. This is reflected in sector-specific legislation such as Federal Decree-Law No. 49 of 2023 Regulating the Use of the Human Genome, alongside the broader personal data protection regime.
In practice, this federal framework is implemented through a layered system of regulators and ethics bodies at Emirate and institutional level, including within specialist healthcare free zones. As a result, governance, ethics approval and data access requirements may vary depending on where a project is anchored and which public healthcare entities are involved. These layers are particularly relevant for AI-enabled genomics projects, where tools may evolve from research analytics into clinical decision support over time. In such cases, regulatory classification, ethics oversight and data governance tend to converge, and misalignment between them can create friction just as projects move from pilot to scale.
A practical consequence is that informed consent is not a procedural formality. The scope of consent directly affects whether genomic data can be reused, combined with new analytical tools or applied to clinical contexts beyond those originally envisaged. Secondary use therefore needs to be designed into projects from the outset and not addressed as a compliance exercise once datasets already exist.
Data localisation and cross-border use present further constraints. While international collaboration is possible, genomic datasets generated within national programmes are often structured to remain under local control, subject to applicable approvals and governance frameworks. In practice, this may limit outright data transfer and require localised analysis models or controlled remote access instead. Similar considerations apply to biological samples, where export for sequencing or further analysis may be restricted even where digital processing is permitted.
In practice, ownership of AI-generated outputs becomes a pressure point. Models trained on local genomic data can generate insights of global relevance, but ownership of those insights does not automatically follow ownership of the model. Rights depend heavily on how access, purpose and permitted use were structured at the outset, particularly where public healthcare bodies or national programmes are involved.
A defining feature of the UAE approach is that population-scale genomic data is treated as a national health asset, with the state acting as custodian and gatekeeper of access and use. Data generated under national programmes remains under state control, with access granted on a permission-based and purpose-limited basis. Participation in these programmes does not confer ownership of the underlying dataset. Rights to reuse, combine or commercialise insights derived from the data depend on the terms of access, the scope of consent and the governance framework put in place at the outset. Legal structuring therefore forms part of system design rather than a downstream consideration.
What this means in practice for organisations engaging with the UAE
For pharmaceutical companies, medtech firms and AI developers, these issues shape how UAE projects should be structured from day one.
Contracting needs to address more than access to data and headline IP ownership. Distinctions between raw data, derived data and model outputs matter in practice, particularly where reuse, onward deployment or integration into global platforms is anticipated. Where public healthcare systems are involved, long-term governance expectations are embedded and difficult to revisit once data generation has begun.
Regulatory pathways also require early attention. AI tools that influence diagnosis, treatment or clinical decision-making may fall within medical device or diagnostics regimes, while their underlying data use remains subject to genomic-specific controls. Treating regulatory classification as a late-stage issue commonly results in friction at precisely the point when scaling becomes commercially critical.
What WHX signals about the direction of travel
The prominence of genomics and AI at WHX reflects a broader shift in the UAE healthcare landscape. These technologies are being integrated on the assumption that they will operate at population scale and over long time horizons.
For organisations engaging with the UAE, this means legal strategy is part of system design rather than a downstream compliance function. Projects that are structured with consent, governance, data control and regulatory classification in mind are more likely to scale and integrate into global portfolios. Without that early legal and governance structuring, technical success in the UAE can be difficult to convert into scalable, transferable value.
