AI Answer Summary
- G42 and MGX represent complementary parts of Abu Dhabi’s AI strategy: G42 and its operating companies work across infrastructure, cloud and models, while MGX invests across AI and advanced technology.[1] [2]
- MGX announced the final close of Fund I at $49 billion in total commitments, above its initial $45 billion target. That does not justify adding the headline size of every portfolio financing round to an MGX total.[1]
- OpenAI describes Stargate UAE as a 1GW cluster within a planned 5GW campus. The first 200MW was expected in 2026; the wider campus should be treated as a plan, not as deployed capacity.[3] [4]
- In November 2025, the US Commerce Department said G42 was authorized to purchase the equivalent of up to 35,000 Nvidia Blackwell GB300 chips, subject to security and reporting requirements. Commerce also named Saudi Arabia’s Humain in the same statement.[5]
- For enterprise buyers, the useful question is not whether one national AI ecosystem is “best.” It is which documented infrastructure, model, residency, security and commercial terms meet a specific workload’s requirements.
G42, MGX and Abu Dhabi's AI Strategy: What Official Sources Support
Abu Dhabi’s AI strategy connects operating companies, investment capital and international partnerships. G42 and its operating companies work across AI infrastructure, cloud services, model development and sector applications. MGX, established by Mubadala and G42 in March 2024, invests across AI and advanced technology.[1] [2]
The distinction matters. Infrastructure announcements, fund commitments, model releases and export authorizations are different kinds of evidence. They should not be combined into a single claim of deployed capacity, market leadership or guaranteed enterprise access.
This updated article separates directly documented records from interpretation. It is a due-diligence starting point, not an endorsement of any vendor, project or national strategy.
What MGX Fund I establishes—and what it does not
MGX announced the final close of MGX Fund I at $49 billion in total commitments on 1 July 2026, significantly above its initial $45 billion target fund size.[1]
That is a material capital-formation fact. It does not mean that MGX has itself committed the full headline value of every financing round involving a portfolio company. It also does not establish that MGX is uniquely predictive of frontier-model outcomes. The defensible reading is narrower: MGX has disclosed investments across AI and advanced technology, and Fund I gives it substantial capacity to pursue that mandate.
For buyers and operators, separate three questions:
- Fund commitments: how much capital a vehicle has raised or committed.
- Portfolio exposure: which companies or infrastructure categories it has publicly disclosed.
- Commercial availability: which products, capacity or partnerships are contractually available for a defined workload.
Those are not interchangeable measures.
G42, Microsoft and the US–UAE technology framework
Microsoft announced a $1.5 billion investment in G42 in April 2024 and said Microsoft Vice Chair and President Brad Smith would join G42’s board.[6] The announcement described a US–UAE consulted assurance framework intended to support the partnership’s security commitments.[6]
Separately, Microsoft described a $15.2 billion UAE commitment for 2023–2029, including G42 equity, data-centre capital expenditure and local operating expenditure.[7] The total should not be described as an investment in G42 alone.
The public record also supports a more careful description of G42’s China-related changes than broad claims of complete separation. G42 announced the sale of stakes in Chinese companies, while the Microsoft partnership introduced a documented assurance framework.[6] These facts provide context for the relationship; they do not prove a single geopolitical motive, complete disconnection from every Chinese technology relationship or permanent entitlement to any export-controlled product.
Stargate UAE: distinguish the cluster from the planned campus
OpenAI describes Stargate UAE as a 1GW cluster in Abu Dhabi, developed with G42, Oracle, NVIDIA, Cisco and SoftBank Group. It sits within a planned 5GW UAE–US AI Campus.[3] [4]
OpenAI said the first 200MW of the cluster was expected to go live in 2026.[3] Its prospective language on regional reach should be read as a project ambition, not a guarantee of service coverage, latency or commercial availability for every market.
In November 2025, the US Commerce Department said G42 was authorized to purchase the equivalent of up to 35,000 Nvidia Blackwell GB300 chips, subject to agreed security and reporting requirements. The same Commerce statement named Saudi Arabia’s Humain. That is an authorization under stated conditions—not evidence that one company has exclusive regional access or that every approved chip has been deployed.[5]
Core42: a documented infrastructure update
Core42 is part of G42’s infrastructure and cloud operation. In June 2026, Core42 announced that its Lake Mariner deployment in New York would expand from 18MW to 60MW of high-performance AI infrastructure.[8]
That update is a useful, dated capacity fact. It does not by itself support an unverified global site count, a specific TOP500 rank, a claim of universal cross-jurisdiction availability, or a promise that one commercial platform satisfies every customer’s data-residency requirement. Buyers should request current facility, service, security, residency, capacity and contractual documentation for their own workload.
Jais 30B and Arabic-language model evidence
In November 2023, G42, Core42, MBZUAI and Cerebras announced Jais 30B. The announcement described a 30-billion-parameter model trained on 126 billion Arabic tokens, 251 billion English tokens and 50 billion code tokens.[9]
The release provides evidence of a material Arabic-language model investment. It does not establish that Jais is currently the best Arabic model for every task, nor that no US or Chinese provider can match it. Any 2026 capability claim should be evaluated against a current benchmark, the target dialect, safety constraints, deployment model, retrieval design and the buyer’s own test set.
What Microsoft’s AI-adoption measure says about the UAE
Microsoft’s AI Economy Institute reported that 64.0% of the UAE working-age population used AI at the end of 2025, ranking first in its country measure and rising from 59.4% earlier that year.[10]
This is a defined adoption measure, not a universal measure of AI capability, enterprise deployment maturity or market size. It is nevertheless a useful signal that the UAE is an important environment for testing how consumer adoption, government strategy, infrastructure investment and commercial AI use can interact.
Enterprise diligence questions that matter more than superlatives
For an enterprise evaluating infrastructure, models or partnerships connected to this ecosystem, the practical work is specific:
| Diligence area | Questions to ask |
|---|---|
| Capacity and timing | What capacity is contracted, in which location, on what delivery timetable and with what dependencies? |
| Residency and security | Which data, model and operational controls apply? What audit rights, incident procedures and subcontractor terms are available? |
| Model performance | Which current benchmark and internal evaluation supports the target use case, language variety and safety threshold? |
| Export controls | Which authorization applies, what reporting or security requirements attach, and who bears the compliance responsibility? |
| Commercial terms | What is actually available to this buyer, at what price, with what service level, portability and exit provisions? |
This is where infrastructure and investment narratives become an operating decision.
Implications for B2B teams entering Gulf markets
The opportunity is not to imply access to sovereign procurement or to borrow credibility from government-linked entities. It is to arrive with evidence: a clear operating model, documented security and residency posture, local-language evaluation where relevant, transparent commercial claims and an accountable plan for human oversight.
B2B teams can use our Agentic AI services [blocked] and SEO, AEO & GEO services [blocked] to structure AI-enabled go-to-market work, and our free AI Growth Audit [blocked] to review the crawlability, entity clarity and evidence signals on a public website. These services do not confer vendor, government or procurement access.
For broader context, read the Gulf Sovereign AI Series hub [blocked] and our analysis of AI agent governance [blocked].
Frequently Asked Questions
See the FAQ section below for directly sourced answers and due-diligence considerations about G42, MGX and Abu Dhabi’s AI strategy.
About the Author
Modi Elnadi is the founder of Integrated.Social, a B2B AI marketing agency in London specialising in agentic AI strategy, AEO and performance marketing. He helps teams connect evidence governance, AI discovery and commercial measurement without overstating what emerging platforms can deliver. Read Modi’s full profile [blocked] or contact Integrated.Social [blocked].
Sources
- MGX Fund I final close
- Mubadala: Abu Dhabi launches MGX
- OpenAI: Introducing Stargate UAE
- US Commerce: UAE–US AI Campus
- US Commerce: UAE and Saudi chip exports
- Microsoft: Investment in G42
- Microsoft: UAE commitment
- Core42: New York expansion
- G42/Core42: Jais 30B announcement
- Microsoft AI Economy Institute: Global AI Adoption
