AI Investments in the Gulf: Opportunities and Surveillance Risks

AI Investments in the Gulf: Opportunities and Surveillance Risks is a long-form research piece by Metehan Durmaz, SMEX's Policy Analyst, published on 19 May 2025 on the organisation's research line. The report maps the Gulf Cooperation Council states' simultaneous emergence as a leading global destination for AI infrastructure investment and as a region whose documented surveillance-against-civil-society record extends into the AI-enabled tooling that the new investment is putting in place — covering Saudi Arabia, the United Arab Emirates, Qatar, Oman, and Bahrain — and is structured around four named sections: Key Investments, Key Stakeholders, Local Investment Benefits, and Potential Threats and Digital Rights Implications.

Argument and methodology

The report's central thesis is that AI policy in the Gulf cannot be analysed separately from the regional political economy of surveillance — that the same regimes building sovereign-AI infrastructure through Saudi Vision 2030 and the UAE National AI Strategy 2031 are the regimes whose use of imported surveillance tooling against journalists, activists, and human-rights defenders is already publicly documented, and that the consolidation of advanced AI infrastructure in their hands risks automating and scaling that surveillance rather than constraining it. The methodology, as the report itself states, is desk research drawing on publicly available information — online articles, research papers, and investment documentation — to map who is investing, what is being built, and where the resulting infrastructure intersects with the existing surveillance record.

The first half maps the investment landscape sector by sector. The report names the Stargate Project — the $500 billion United States AI-infrastructure initiative whose financing partly routes through MGX, the Abu Dhabi state-backed investment vehicle launched in 2024 — alongside Microsoft's $1.5 billion investment in G42, the Abu Dhabi AI firm, as the centre of gravity of US–UAE AI cooperation; the $14.9 billion in AI investment commitments announced at LEAP 2025 in Saudi Arabia and the $40 billion Public Investment Fund commitment dedicated to AI initiatives under the SDAIA umbrella; the Saudi-side partnerships with OpenAI, IBM, Cerebras Systems, Oracle, SoftBank, and Databricks on the US-aligned side and with Alibaba (the $238 million cloud-computing deal with Saudi Telecom), Huawei, Baidu, Tencent Cloud's Saudi-anchored Middle East region, and the Chinese generative-AI firm Zhipu AI on the China-aligned side, framing the Gulf as the principal site of the US–China AI-infrastructure competition outside the two main powers themselves; the $500 billion NEOM smart-city project and NEOM's $100 million investment in the Chinese autonomous-vehicle firm Pony.ai; and Qatar's $250 million investment in Builder.ai alongside the Qatar Investment Authority and Oman Investment Authority as additional state-investment anchors. The corollary figure — that private-sector AI investment in the Middle East stood at only about $700 million as of November 2024 — is named to establish that the AI infrastructure being built in the region is overwhelmingly state-led and sovereign-wealth-fund-led, not market-led, and therefore answers to the political economy of the states funding it rather than to private-sector commercial pressure.

The second half traces the surveillance-and-rights register onto the same infrastructure. The historical anchor is Pegasus spyware, the NSO Group product whose documented deployment against Saudi women's-rights activist Loujain al-Hathloul and Emirati human-rights defender Ahmed Mansoor the report cites as the established pattern of regional surveillance abuse that the new AI investment now scales rather than displaces. The Oyoon programme — the Dubai Police AI-integrated facial-recognition and behavioural-analytics system whose 300,000-plus-camera footprint the report identifies — is named as the regional smart-city counterpart to the Xinjiang and West Bank ethnic-monitoring systems whose comparative-modelling shape the report draws on for analytic context. ToTok, the messaging application exposed as a mass-surveillance instrument allegedly used by UAE authorities, is named as the regional pattern of state-aligned consumer applications carrying surveillance infrastructure inside ordinary digital products. And the broader argument — that AI foundation models can be shaped at training time to suppress dissent or erase state-critical narratives, that chip and infrastructure control gives authoritarian entities "unprecedented visibility over populations," and that the absence of robust regulatory frameworks and independent oversight across the Gulf means the constraints on these uses are weak by design — anchors the report's case for treating Gulf AI investment as one strand of a wider political-economy-of-repression analysis rather than as a separate AI-policy domain.

Recommendations and posture

The report does not issue prescriptive policy recommendations. Its framing is that the issue requires "transparent governance, stringent regulation, and active engagement from both civil society and international stakeholders" — and that the prerequisite for any of those is sustained civil-society vigilance over who the central actors and stakeholders behind the surge in Gulf AI investment actually are. The local-benefits section of the report names a counter-thread: locally developed AI models that better serve Arabic dialects, Farsi, Turkish, and Kurdish, and that address the systemic content-moderation failures that platforms like Meta exhibit in the region, offer a substantive opportunity that the same investment surge could be steered toward — but only if the civil-society and international-stakeholder safeguards the report calls for are in place. The piece is, in posture, a cautionary mapping exercise rather than an advocacy document — produced for SMEX's civil-society-evidence-base register, intended as a baseline reference document for regional and international advocacy, journalism, and policy work on AI-and-human-rights in the Gulf rather than as a coalition-organising artefact.

Posture within the corpus

Within the corpus, AI Investments in the Gulf: Opportunities and Surveillance Risks is the first publication anchored on the West Asia and North Africa region and the first Publication co-produced under SMEX's research-publications imprint. The publications slate has previously clustered around the United States (Unmasking AI, Bug Bounties for Algorithmic Harms, Comply To Fly?, On the Dangers of Stochastic Parrots, A Hazard to Human Rights, Losing Humanity), the United Kingdom (Participatory Data Stewardship, Not a drop to drink), Continental Europe (Automating Society Report 2020, Civil Society Statement on the EU AI Act), Latin America (Decisiones automatizadas en la función pública en América Latina), South Asia ("No Internet Means No Work, No Pay, No Food"), and East and Southern Africa (Legal Analysis of Laws, Policies and Government Strategies Relating to Artificial Intelligence in Kenya, Mauritius, Rwanda, South Sudan, Tanzania, Uganda and Zambia) — leaving the WANA region without a publication anchor despite SMEX's standing as the corpus's principal Arabic-language WANA digital-rights organisation. The Durmaz report installs the first Arabic-language civil-society analytic on Gulf AI-investment political economy as the corpus's reference document on the region.

The report exercises the SMEX org-as-publisher pattern for the first time on the corpus's publications slate, and the substantive contribution it carries — that AI policy and surveillance are connected facets of the same political economy of regional repression and not separate policy domains — is the recurring posture of SMEX's 2024-2025 AI-and-surveillance research line, alongside the policy brief "The Human Rights Risks of Big Tech's Cloud Expansion in the Gulf" (5 March 2025), "Confronting Structural Silencing: Challenges and Resistance Among Digital Feminist Activists in Lebanon" (25 April 2025) on Arabic-language content-moderation suppression of feminist activism, and "Click, Load, Kill: A Look into the Cyberweapon Industry in the WANA Region" (1 October 2025) on Cellebrite, Intellexa, and Pegasus. Within that line, AI Investments in the Gulf anchors the political-economy-of-AI-investment analytic specifically — distinct from the cloud-expansion brief's infrastructure-procurement frame, from the cyberweapon-industry report's surveillance-supply-chain frame, and from the platform-policy frame the Durmaz-authored "Meta AI's darker side" (3 July 2025) and the Cupler-authored "A Brief Overview of AI Use in WANA" (26 May 2023) carry on the consumer-AI side.

In the corpus's wider regional shape, the report is the structural counterpart of the Derechos Digitales-led Decisiones automatizadas en la función pública en América Latina on the Latin American side and the Paradigm Initiative-led Legal Analysis of Laws, Policies and Government Strategies Relating to Artificial Intelligence in Kenya, Mauritius, Rwanda, South Sudan, Tanzania, Uganda and Zambia on the African side: three Global Majority civil-society anchors, each producing a region-specific AI-and-rights evidence base under its own institutional imprint, with SMEX the Arabic-language WANA pole inside the same Global Network for Social Justice and Digital Resilience infrastructure that connects all three. The report is referenced from the SMEX body's AI-and-surveillance research section as the principal regional civil-society analytic on Gulf AI investment, and the substantive frame it carries — that the AI infrastructure surge is one strand of a wider political economy of regional repression and must be analysed alongside the existing surveillance record rather than as a separable AI-policy domain — is the standing reference posture the corpus carries on SMEX's Gulf-AI analytic line.