AI Is Not a Product. It Is Infrastructure. And Infrastructure Is a Right.

There is a useful thought experiment for understanding where we are with artificial intelligence. Imagine a city that decides to run its water supply, its hospitals, its schools, its courts, and its social welfare system through a single private company — a company headquartered in another country, accountable to its shareholders, and operating under terms of service rather than public law. Most governments would consider this arrangement incompatible with the basic obligations of the state to its citizens. The infrastructure is too important. The accountability is too diffuse. The power is too concentrated.

Now consider what the UAE announced on June 14, 2026: a national mandate to run half of all government services through agentic AI within two years. Or Saudi Arabia's Project Transcendence — $100 billion in sovereign AI infrastructure touching healthcare, education, finance, and public administration. Or Qatar's binding AI governance framework for its financial sector. The Gulf is not experimenting with AI. It is embedding it into the architecture of the state.

When AI mediates access to the services that define a citizen's relationship with their government — healthcare decisions, benefit eligibility, permit approvals, judicial processes, educational placement — it is no longer a product. It is infrastructure. And infrastructure, at that scale, carries human rights obligations.

The global community has been arriving at this conclusion with increasing urgency. The question for the Gulf — and for the AI companies that want to serve it — is whether they are building accordingly.

The international consensus is not theoretical

In November 2021, all 193 member states of UNESCO adopted the world's first global standard on AI ethics. ^[1] The Recommendation on the Ethics of Artificial Intelligence is unambiguous at its core: AI must respect human rights and human dignity. It establishes that AI systems which affect access to public services, healthcare, education, and economic opportunity must be designed with fairness, transparency, human oversight, and accountability as non-negotiable requirements — not optional features.

In September 2024, the Council of Europe opened for signature the Framework Convention on Artificial Intelligence — the first legally binding international treaty on AI, signed by the United States, the European Union, the United Kingdom, and more than fifty countries. ^[2] It establishes, in enforceable international law, that activities within the lifecycle of AI systems must be fully consistent with human rights, democracy, and the rule of law. It requires signatories to provide effective procedural guarantees to people affected by AI decisions, and explicitly contemplates banning AI applications whose risks are incompatible with human rights standards.

The UN Secretary-General's High-Level Advisory Body on Artificial Intelligence, in its September 2024 report "Governing AI for Humanity," identified what may be the most consequential finding in the entire global AI governance debate: 118 countries are currently excluded from the major AI governance initiatives that will shape the standards, norms, and rules governing this technology. ^[3] The report calls explicitly for AI governance anchored in the UN Charter and international human rights law — and warns that without equitable access, AI will deepen rather than reduce global inequality.

And then, in May 2026, Pope Leo XIV issued Magnifica Humanitas — the first papal encyclical dedicated entirely to artificial intelligence. He warned that entrusting irreversible decisions to AI systems is "not permissible." He declared that "a more moral AI is not enough if that morality is determined by a few." ^[4] In a single sentence, he articulated the human rights argument more precisely than most policy documents have managed: the problem is not just whether AI is ethical, but who determines what ethical means.

The Gulf is not peripheral to this conversation — it is central to it

There is a tendency in global AI governance discussions to treat the Gulf as a recipient of norms developed elsewhere. This is analytically wrong and strategically consequential for any company trying to understand the market.

The UAE, Saudi Arabia, and Qatar are not implementing AI in the way a typical enterprise deploys a software tool. They are deploying AI at sovereign scale — into the fundamental services of the state — on timelines and at levels of ambition that most Western governments have not approached. The UAE's AIDA mandate is not a pilot program. Saudi Arabia's Year of AI is not a branding exercise. These are structural commitments.

This creates a distinctive human rights dynamic. In societies where government and citizen are bound by deep obligations of trust, where religion and governance are intertwined, and where the state's legitimacy rests in part on its provision of welfare and service — the deployment of AI into those systems is not a neutral technology decision. It is a social contract question.

The Soufan Center, in its April 2025 analysis of Gulf AI strategy, explicitly raised the intersection of AI ambition and civil liberties — noting that the scale of AI deployment in the UAE and Saudi Arabia introduces questions around surveillance, data sovereignty, and algorithmic accountability that the region's governance frameworks are still developing. ^[5] The Crowell and Moring assessment of Middle East AI and data security frameworks noted that Gulf states are actively working to align their data protection architectures with global standards — including hosting the Global Privacy Assembly in Dubai in late 2026 — precisely because they understand that international credibility requires it. ^[6]

What this means for AI vendors operating in the region

The human rights frame changes the vendor selection calculus in ways that are not yet fully priced in by most enterprise AI companies.

A government that is building AI into the delivery of healthcare decisions, educational placements, and social welfare eligibility is not just buying a technology product. It is, in effect, delegating part of its social contract obligations to the AI system — and by extension, to the company that built it. That delegation carries accountability. It carries auditability requirements. It carries the need for human oversight mechanisms. It carries the obligation to explain outcomes to affected citizens.

These are not abstract requirements. The UNESCO framework, the Council of Europe convention, and the UN Advisory Body report all converge on the same operational requirements: transparency, human oversight, non-discrimination, risk assessment, and the ability to challenge AI-driven decisions. Gulf governments building at sovereign scale will — as their regulatory frameworks mature — need to demonstrate that the AI systems embedded in their public services meet these standards. The vendors they select will need to meet them too.

This is where Anthropic's Constitutional AI framework becomes strategically relevant in a way that goes beyond marketing. ^[7] Constitutional AI is not a brand position. It is a technical and governance architecture — a structured approach to ensuring AI systems operate within defined value boundaries, with interpretable reasoning and human oversight built into the model rather than bolted on afterward. It is, in substance, the kind of approach that the UNESCO recommendation, the Council of Europe convention, and the UN report are calling for. Not because Anthropic designed it to match those frameworks, but because it was designed to solve the same underlying problem.

The Gulf angle: rights, values, and the legitimacy of the state

There is a dimension of this argument that is specific to the Gulf and rarely articulated in Western AI governance discussions.

Gulf governments do not operate on the assumption of a separation between state authority and moral framework. In the UAE, Saudi Arabia, Qatar, and Bahrain, governance is explicitly grounded in Islamic values — justice, equity, the protection of the vulnerable, the accountability of the powerful. These are not rhetorical commitments. They are embedded in the constitutional and legal architecture of each state.

When Sheikh Mohammed bin Rashid announced AIDA with the framing of "a government built around people, not paperwork," ^[8] he was making a rights-adjacent statement — the explicit commitment that the technology serves the citizen, not the bureaucracy. When Saudi Arabia's AI ethics principles include human oversight and accountability as mandatory pillars, they are articulating values that align precisely with the international human rights framework. When Qatar's Central Bank requires board-level accountability for AI outcomes in financial services, it is operationalizing the same accountability obligations the Council of Europe treaty requires.

This convergence is not coincidental. It reflects something important about how Gulf governments understand the legitimacy of the state: it rests, in part, on the demonstrable protection of citizens' interests. An AI system that makes opaque decisions affecting people's access to healthcare or housing is not just a technology risk — it is a legitimacy risk for the government that deployed it.

The argument this makes for a different kind of AI partner

At Qlabs Intelligence, we have argued across this series that the Gulf requires AI partners who understand the institutional landscape, not just the technology. This article adds a third dimension to that argument.

The Gulf is building AI into the architecture of the state, at a moment when the international community has established that AI-at-scale carries human rights obligations, and at a moment when the values embedded in Gulf governance are unusually well-aligned with the accountability requirements those obligations demand.

The AI companies that will succeed in this market over the long term are not the ones with the most aggressive sales function or the largest compute cluster. They are the ones whose governance architecture — whose approach to transparency, accountability, human oversight, and the protection of the people affected by their systems — is compatible with what Gulf governments need to fulfill their obligations to their citizens.

That is a structural argument, not a sentimental one. And it points, with unusual precision, to a short list of companies.