Middle East AI Actors

Quick Assessment

Overview

The Middle East has emerged as a significant arena of AI investment, infrastructure development, and geopolitical competition. Driven primarily by the Gulf states' sovereign wealth and their ambitions to diversify beyond hydrocarbons, countries including the UAE, Saudi Arabia, and Qatar have committed tens of billions of dollars to national AI strategies, data center construction, and partnerships with leading US technology firms. Simultaneously, Israel maintains a distinct profile anchored in military AI, elite technical talent, and a dense startup ecosystem. Smaller regional players including Turkey, Egypt, and Jordan are beginning to position themselves as secondary AI hubs, often supported by Gulf capital.

The region's AI actors operate at the intersection of economic transformation and geopolitical tension. Gulf states occupy a structurally unusual position — capital-rich but compute-hungry — and have had to negotiate carefully with the United States over access to advanced semiconductors, while simultaneously maintaining relationships with Chinese technology partners. This "balancing act" has attracted scrutiny from US policymakers and AI safety researchers who worry about advanced AI capabilities diffusing to autocratic governments with opaque governance structures and documented records of using technology for surveillance and repression.

The AI safety and effective altruism communities have engaged with Middle East AI actors primarily through the lens of power concentration risk and export-control policy. The region's combination of vast capital, rapid infrastructure buildout, and limited domestic accountability mechanisms makes it a salient case study for questions about who gains access to frontier AI capabilities and under what conditions.

History

Origins: National Strategies and Early Moves (2017–2022)

The contemporary wave of Middle East AI development traces to roughly 2017–2019, when Gulf governments began formalizing AI as a pillar of economic diversification. The UAE established the world's first Minister of State for Artificial Intelligence, appointing Omar Al Olama to that role, and launched its National AI Strategy 2031.¹ Saudi Arabia embedded AI in its Vision 2030 framework and established SDAIA (the Saudi Data and AI Authority) by royal decree in 2020 to regulate and coordinate national AI development.² Qatar published its National Artificial Intelligence Strategy in 2019.³

These early strategies were largely aspirational, focused on attracting foreign talent and signing partnership agreements with global technology firms. The UAE's G42, founded in Abu Dhabi, began assembling a portfolio of AI and cloud computing ventures during this period, positioning itself as the region's flagship AI conglomerate. Israel, drawing on decades of military intelligence technology and its Unit 8200 talent pipeline, was already producing globally competitive AI startups well before the Gulf states entered the field.

Infrastructure Buildout and the Compute Race (2022–2024)

The release of large language models and the generative AI boom substantially accelerated Gulf ambitions. Gulf states recognized that AI leadership requires not just software and talent but massive compute infrastructure — data centers, GPUs, and high-bandwidth networking — which required navigating US export-control regimes.

The UAE's G42 became a central figure in US-UAE semiconductor diplomacy. The US Commerce Department approved the export of 70,000 advanced Nvidia chips to the UAE, conditioned on G42 distancing itself from Chinese technology firms.⁴ This represented a significant concession by US authorities and a signal that the UAE was being treated as a preferred partner — albeit a monitored one — in the global AI infrastructure race.

Saudi Arabia's Public Investment Fund (PIF) announced a $40 billion commitment to AI in 2024, spanning semiconductors, data centers, and direct investments in AI companies.⁵ The kingdom also hosted the third Global AI Summit in Riyadh in September 2024 and partnered with the OECD for an AI Policy Observatory the same month.⁵

2025: Sovereign AI and the HUMAIN Moment

Saudi Arabia's most consequential AI initiative launched in 2025: HUMAIN, a $15 billion sovereign AI project backed by PIF and associated with Crown Prince Mohammed bin Salman. In May 2025, HUMAIN signed a memorandum of understanding with Qualcomm covering advanced AI data centers, edge devices, cloud infrastructure, and a semiconductor design center.⁵ The launch of HUMAIN signaled Saudi Arabia's intent to move beyond passive investment into building sovereign AI infrastructure and indigenous model capabilities.

Key Activities

United Arab Emirates

G42 (Group 42 Holding)

G42 is the UAE's leading AI and cloud conglomerate, headquartered in Abu Dhabi. Led by CEO Peng Xiao, G42 has built a diversified portfolio spanning healthcare AI (through its M42 subsidiary), space technology (Space42), and large-scale compute infrastructure. The company has made notable investments in OpenAIOrganizationOpenAIComprehensive organizational profile of OpenAI documenting evolution from 2015 non-profit to Public Benefit Corporation, with detailed analysis of governance crisis, 2024-2025 ownership restructuri...Quality: 62/100 and xAI, and is driving the development of a 5-gigawatt AI campus equipped with Cerebras supercomputers.⁶ G42's partnership with Microsoft — which includes equity investment by Microsoft — anchors it firmly in the US technology ecosystem, a relationship that has helped it navigate export-control requirements.⁴

G42 partners with major US hyperscalers including Amazon Web Services, Microsoft Azure, Google Cloud, and Oracle Cloud for data center operations.⁴ The chip-approval conditions requiring distance from Chinese technology firms have made G42's strategic alignment with the US explicit and consequential.

Technology Innovation Institute (TII) and the Falcon Models

The Technology Innovation Institute, an Abu Dhabi research center operating under the Advanced Technology Research Council (ATRC), is the origin of the Falcon family of large language models. Falcon models have been released as open-weight models and have ranked among the most capable openly available LLMs. TII represents the UAE's attempt to build genuine indigenous AI research capacity, rather than relying solely on partnerships and procurement.

Mohamed bin Zayed University of Artificial Intelligence (MBZUAI)

MBZUAI is the world's first graduate-level university dedicated to AI, located in Abu Dhabi. It functions as the academic anchor of the UAE's AI ecosystem, training researchers and hosting international faculty. MBZUAI's existence reflects the UAE's understanding that sustainable AI leadership requires human capital development, not just infrastructure.

Presidential Court AI Office and Omar Al Olama

Omar Al Olama, as Minister of State for Artificial Intelligence, has served as the public face of UAE AI diplomacy and strategy. His office coordinates AI policy across government and represents the UAE in international AI governance discussions. The UAE has used this institutional structure to position itself as a legitimate voice in global AI governance conversations, including advocating for positions on export controls that reflect its interests as a capital-rich but chip-constrained actor.

Mubadala Investment Company

Mubadala, Abu Dhabi's sovereign wealth fund, holds significant stakes in AI-adjacent technology companies and has been an active LP in global venture funds with AI exposure. Mubadala's investments complement G42's operational role and ATRC's research mandate, providing financial infrastructure for the broader UAE AI ecosystem.

Saudi Arabia

HUMAIN and PIF

HUMAIN is Saudi Arabia's flagship sovereign AI initiative, launched in 2025 under PIF backing with a stated capitalization of $15 billion.⁵ It represents Saudi Arabia's most direct attempt to build nationally controlled AI infrastructure and capabilities. The Qualcomm partnership announced in May 2025 suggests an intent to develop Saudi semiconductor design capabilities alongside data center operations — an ambition that, if realized, would significantly reduce dependence on foreign chip suppliers.

SDAIA and ALLaM

SDAIA developed ALLaM, described by Saudi authorities as "sovereign AI" built by Saudi engineers for Arabic-speaking users. According to SDAIA's public communications, ALLaM was trained on a 500-billion-token Arabic dataset — assembled by coordinating 16 government entities — making it one of the largest Arabic-language training corpora assembled for a single model.² More than 400 subject matter experts tested the model through over a million prompts before deployment. The model explicitly encodes Islamic values and regional cultural context and is deployed through HUMAIN Chat. These claims about scale and cultural alignment come from SDAIA itself and should be understood as official characterizations.

KAUST

King Abdullah University of Science and Technology (KAUST) serves as Saudi Arabia's primary research university with serious AI and machine learning programs. KAUST's international faculty and research partnerships provide human capital that domestic institutions alone cannot yet supply.

Vision 2030 and NEOM

Saudi Arabia's Vision 2030 economic diversification strategy frames AI as central to reducing oil dependence. The NEOM megaproject — including the controversial "The Line" development — has been marketed with AI-driven smart city ambitions, including AI-managed infrastructure and logistics. Whether these ambitions are achievable at the stated scale remains contested.

Israel

Unit 8200 Pipeline

Israel's military intelligence Unit 8200 has functioned as a talent forge for the global AI and cybersecurity industry. Alumni of Unit 8200 have founded or occupied senior roles at frontier AI labs including OpenAIOrganizationOpenAIComprehensive organizational profile of OpenAI documenting evolution from 2015 non-profit to Public Benefit Corporation, with detailed analysis of governance crisis, 2024-2025 ownership restructuri...Quality: 62/100, AnthropicOrganizationAnthropicComprehensive reference page on Anthropic covering financials ($380B valuation, $14B ARR at Series G growing to $19B by March 2026), safety research (Constitutional AI, mechanistic interpretability...Quality: 74/100, and Meta AI. This pipeline has given Israel disproportionate influence in the development of frontier AI systems relative to its size.

AI Startups

Israel hosts a dense cluster of AI companies with global reach. AI21 Labs develops large language models and AI writing tools. Lightricks focuses on AI-powered creative tools and image/video generation. D-ID specializes in AI-generated video from still images, including digital human synthesis. These companies represent commercialization of Israel's deep technical talent pool.

IDF AI/ML Applications

The Israel Defense Forces have been among the most active military adopters of AI for operational purposes. Israel used AI-enabled intelligence processing and targeting systems during the Unity Intifada in 2021, which Israeli commentators described as the "world's first AI war."⁷ Following the October 7, 2023 Hamas attacks, Israel deployed AI targeting systems including reported programs named "Lavender" and "Wolfpack" at significantly larger scale in Gaza and the West Bank.⁸ These deployments have generated substantial criticism from human rights organizations and AI ethics researchers (see Criticisms section).

Academic Research: Weizmann Institute and Technion

The Weizmann Institute of Science and the Technion (Israel Institute of Technology) maintain internationally recognized AI and machine learning research programs, providing the academic foundation for Israel's commercial and military AI ecosystem. The Israel Innovation Authority (IIA) has also run dedicated AI programs to support domestic AI development.

Qatar

Qatar Computing Research Institute (QCRI)

QCRI, part of Hamad Bin Khalifa University, is Qatar's primary AI research institution. It focuses on areas including Arabic language processing, social computing, and data analytics. Qatar's "Al-Fanar" project, launched in May 2024 with a $2.5 billion government commitment to digital transformation, aims to develop Arabic-language AI tools.⁵

Qatar's AI footprint is smaller than those of the UAE and Saudi Arabia but benefits from similarly vast sovereign wealth. The Qatar Foundation and associated research institutions have received significant funding for multidisciplinary AI research, including healthcare AI governance frameworks.⁹

Turkey and Egypt

Turkey and Egypt occupy emerging positions in the regional AI landscape. Egypt launched an AI Council and National AI Strategy in 2019 and issued an Egyptian Charter for Responsible AI — a regional first in AI governance and ethics — and has invested in hyperscale data centers, ranking in the top ten regionally on government AI indices.¹⁰ Jordan and Morocco are launching their own AI initiatives with Gulf investor support, expanding the Gulf-centered AI ecosystem beyond the GCC. Turkey has ambitions in AI but faces significant economic constraints relative to Gulf states.

Funding

Gulf AI actors are distinguished globally by access to sovereign wealth capital at a scale that few other regional ecosystems can match.

MENA technology spending is projected to reach $169 billion by 2026.⁵ The generative AI opportunity in the GCC has been estimated at $21–35 billion from near-term use cases.¹² Top venture investors active in the region include GAIA AI Accelerator, Flat6Labs, 500 Global, Wa'ed Ventures, and G42.¹¹

Key People

Helen Toner, interim executive director of the Center for Security and Emerging Technology (CSET) and former OpenAIOrganizationOpenAIComprehensive organizational profile of OpenAI documenting evolution from 2015 non-profit to Public Benefit Corporation, with detailed analysis of governance crisis, 2024-2025 ownership restructuri...Quality: 62/100 board member, has analyzed US Gulf AI policy specifically. In discussions on the EA Forum, she characterized US approval of supercomputer sales to Gulf autocracies as raising serious questions, rejecting the argument that "if we don't sell it, China will" on the grounds that Chinese production constraints limit this scenario.¹⁴ She highlighted that Gulf rulers maintain close ties with Chinese leadership and participate in joint military exercises with China, complicating the framing of Gulf states as straightforwardly US-aligned partners.¹⁴

Criticisms

Surveillance, Repression, and Authoritarian Enablement

The most consistent criticism of Middle East AI actors — particularly UAE and Saudi Arabia — concerns the dual-use nature of their AI investments. US firms including Microsoft and Google have supplied facial recognition, predictive analytics, and other AI tools to Gulf governments, which have deployed these systems not only for public services but for monitoring activists, journalists, and dissidents.¹⁵ Critics note that these exports frequently occur without meaningful human rights oversight, and that AI systems trained primarily on non-Arabic data exhibit higher error rates for darker skin tones, disproportionately affecting the populations they are applied to.¹⁵ In Lebanon and Jordan, AI-based classification systems have reportedly sorted refugees by sectarian background, affecting access to resources.¹⁵

Military AI and Civilian Harm

Israel's use of AI targeting systems in Gaza — including reported programs named "Lavender" (for target generation) and "Wolfpack" (for West Bank tracking operations) — has attracted significant criticism from human rights organizations. Critics argue these systems scale biases present in training data, compress decision timelines in ways that increase civilian casualty risk, and diffuse moral and legal accountability for targeting decisions.⁸ The Stop Killer Robots campaign, launched in 2013, has used documentation of AI weapon deployments in the Middle East as a central part of its advocacy. US AI decision-support systems (including Palantir's Project Maven) used in operations against Houthis have also been criticized for contributing to casualties that exceeded prior decades of operations in the region.⁸

Geopolitical Hedging and Export-Control Concerns

The EA and AI safety communities have raised concerns about the concentration of advanced AI capabilities in Gulf autocracies.¹⁴ The tension is between the possibility that fewer AI-capable states might reduce diffusion of dangerous capabilities, and the countervailing risk that concentrating those capabilities in governments with weak accountability structures and demonstrated interest in repression may itself be dangerous. Helen Toner and others have questioned whether US semiconductor export approvals to the UAE reflect coherent strategic goals or primarily commercial pressures.¹⁴

Influence Operations and Disinformation

Iran and non-state actors including ISIS and the Muslim Brotherhood have exploited commercially available AI tools for influence operations. The Iranian group STORM 2035 used ChatGPT to generate content for fake news sites and social media accounts in multiple languages, targeting polarizing issues including the Israel-Hamas conflict and US politics, before being disrupted by OpenAIOrganizationOpenAIComprehensive organizational profile of OpenAI documenting evolution from 2015 non-profit to Public Benefit Corporation, with detailed analysis of governance crisis, 2024-2025 ownership restructuri...Quality: 62/100 in 2024.¹⁶ The Rapid Support Forces (RSF) in Sudan's civil war (which began April 2023) represent one of the first documented cases of a violent non-state actor using generative AI at scale in an active conflict, including AI-generated audio recordings released as early as October 2023.¹⁷ AI-generated deepfake images from Middle East conflicts have reached over a billion views on X, according to some estimates, overwhelming authentic documentation.¹⁶

Arabic AI Monitoring Gaps

AI moderation tools trained on English or Modern Standard Arabic fail systematically on Arabic dialects, code-switching, and emoji-heavy text common in regional social media. This creates both false positives (over-censorship of legitimate content) and false negatives (failure to detect extremist recruitment material), providing a structural advantage to actors who understand these gaps.¹⁵ Jordan and Iraq have initiated projects to train AI on local dialects specifically for abuse detection, with planned expansion to Egypt.¹⁵

Key Uncertainties

• Whether HUMAIN and SDAIA's ALLaM claims about training data scale and sovereign capability reflect genuine technical achievements or primarily serve political communication goals remains difficult to assess from public information.
• The extent to which G42's commitments to distance from Chinese technology are verifiable and durable is unclear.
• Whether Gulf AI investments will produce genuinely competitive indigenous AI capabilities or primarily fund branded partnerships with US hyperscalers is an open question.
• The degree to which Israeli military AI programs comply with international humanitarian law remains actively contested, with limited independent verification available.
• How US export-control policy toward the Gulf will evolve under changing administrations introduces significant uncertainty for the region's compute trajectory.

Sources