GCC AI News — 13 Nov 2026

The Institute of Foundation Models (IFM), established by Abu Dhabi's Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), released PAN 2 on November 11, an open, general-purpose world model for physical intelligence. Given a starting scene and an action written in words, PAN 2 simulates a video of what happens next, grounded in physics, time and space. IFM calls the approach 'simulative reasoning,' letting people and AI agents ask 'what happens if I do this?' and see the outcome before acting. Built on the Generative Latent Prediction architecture pioneered by IFM founder and MBZUAI president Eric Xing, it placed third overall and first among open-weight models on the Physics-IQ Verified leaderboard. IFM is releasing the model weights, full training and inference code and documentation under the Apache 2.0 license.

World models let AI learn and plan without costly, slow or irreversible real-world trials, a foundation for robotics, education and safer autonomous systems. A Gulf university releasing a competitive one fully open, with training code most labs keep private, puts the UAE at the research frontier and hands the global community something concrete to build on.

A robotics team teaching an arm a new task can let PAN 2 simulate the likely outcome of each candidate action first, filtering out moves that would knock something over before the real robot ever touches the object, saving hardware wear and trial time.

HUMAIN CEO Tareq Amin announced on November 11 in Riyadh that the PIF-backed company's full AI stack will be available in December, anchored by what he called the world's largest open-weight model at 3 trillion parameters, deployed inside Saudi Arabia. 'Sovereign infrastructure, deployed here, ready to be consumed by any builder, enterprise, government without any control claim that leaves the country,' Amin said, adding that the software stack, cloud inferencing and infrastructure are all built and operated by Saudis. He said the state-of-the-art model would run at 88% lower cost and token consumption than comparable models from outside the Kingdom. The December rollout sits beneath HUMAIN+, an applied AI education platform launching in the first quarter of 2027 with a Riyadh campus for about 1,500 learners a year and GPU-backed builder labs.

Most of the world rents its AI from a handful of foreign providers. A Gulf company owning the entire stack, from chips and cloud to a frontier open-weight model, and offering it at a fraction of the cost, means Saudi and regional builders can develop on sovereign infrastructure without sending data or dependence abroad.

A Riyadh startup that cannot afford frontier-model API bills or send its data outside the Kingdom could build its product on HUMAIN's low-cost local model and cloud, keeping both its costs and its data inside Saudi Arabia.

Khalifa University, the GSMA and AMD launched an operator evaluation of RFGPT at MWC Doha 2026, announced November 8, opening the Abu Dhabi-developed radio-frequency foundation model to telecom operators worldwide. Built by Khalifa University's Digital Future Institute, RFGPT combines wireless-signal understanding with natural-language interaction, letting users ask questions about observed signals, identify their characteristics and interpret complex scenes containing multiple signals. It recognizes technologies including Wi-Fi, Bluetooth, LTE and 5G New Radio, and can read key 5G parameters that normally require specialist engineers. Hosted on AMD compute and supported by the GSMA's global operator ecosystem, it is offered as a two-week API trial so operators can test it against real network-planning, spectrum-management and interference-analysis tasks.

Interpreting raw radio-frequency data is slow, specialist work that bottlenecks network planning everywhere. A Gulf university putting a working RF language model in front of the global telecom industry, on stage at a major show, is a rare case of a GCC research lab exporting a specialist AI tool to the world rather than importing one.

A telecom operator planning a 5G rollout can request the trial and ask RFGPT, in plain language, to characterize interference in a congested area, getting a reading in minutes instead of waiting for a scarce RF specialist to analyze the spectrogram by hand.

The Saudi Data and AI Authority (SDAIA) released a comprehensive study on the state of AI in the Kingdom on November 11, measuring progress across seven pillars and 15 sub-pillars. Saudi Arabia ranked third globally in AI talent growth rate under the 2026 Stanford HAI Index, third globally and first regionally in developing leading AI models under the 2025 index, and fifth globally in AI sector growth. From 2021 to 2025, over 69,300 technical graduates entered the market, driving 15% annual growth in domestic AI talent density, while 1.2 million people joined the first phase of the national SAMAI upskilling initiative by July 2026. The study also placed the Kingdom first worldwide in women's empowerment in AI, in AI security and privacy, and in digital readiness, with 90% of public entities now maintaining an AI strategy.

Hardware and funding get the headlines, but people decide whether an AI ecosystem lasts. A government body publishing hard numbers on graduates, upskilling reach and women's participation shows the Kingdom treating talent as measurable national infrastructure, and the top-ranked women's-empowerment finish is a genuinely good story for the region.

A multinational weighing where to open a regional AI hub can read SDAIA's talent-density growth and graduate numbers as evidence it could actually staff an engineering team locally, rather than flying in every specialist from abroad.

Ooredoo Qatar and Microsoft announced on November 10 that they are expanding their collaboration to make Microsoft 365 Copilot available to all Ooredoo Qatar employees, extending access beyond the initial pilot groups as part of the operator's Strategic Digital and AI Transformation Programme. CEO Sheikh Ali bin Jabor bin Mohammad Al Thani said the company wants colleagues across the organization to help shape process improvements, since they best know which steps could be simpler and what matters to customers. The programme pairs Ooredoo's connectivity and digital infrastructure with Microsoft's cloud, data and AI capabilities, and the company said it is also preparing to let employees build and use AI agents for specific tasks within its own security and governance controls. The initiative supports Qatar National Vision 2030 and Digital Agenda 2030.

Handing an AI assistant to every employee, not just a tech team, is how AI actually changes an organization's daily work rather than staying a pilot. A major Gulf operator doing it company-wide and inviting staff to redesign processes is a practical template for how large regional employers can move from experimenting with AI to living with it.

An Ooredoo customer-service manager can use Copilot to summarize long policy documents, draft replies and pull together reports in minutes, freeing time for the judgment calls that actually need a person, while the company watches which use cases prove worth scaling.

Beyon Cyber, part of Bahrain's Beyon Group, announced on November 9 that its proprietary Orryx AI platform has achieved SOC 2 attestation, making it the first GCC-developed agentic AI security-operations platform to earn the internationally recognized credential. SOC 2, established by the American Institute of Certified Public Accountants, provides independent assurance about the design of a platform's controls. Developed in-house in Manama, Orryx AI combines automation and agentic AI to investigate, prioritize and respond to cyber threats, reducing alert noise and analyst workload so teams can respond faster and more accurately. CEO Dr Shaikh Khalid bin Daij Al Khalifa said the attestation marks a step in taking Orryx AI to international markets and reflects a commitment to the highest standards of assurance as the company scales. The credential adds to Beyon Cyber's CREST accreditation.

GCC technology companies are often buyers of foreign software; a Bahrain-built AI platform clearing a demanding international assurance standard flips that script. It signals Gulf-origin products can compete on trust and quality abroad, not just at home, which is exactly the export story the region's tech ambitions need.

A bank outside the region evaluating AI tools to help its small security team triage a flood of alerts can treat Orryx's SOC 2 attestation as independent evidence the platform's controls are sound, lowering the hurdle to trusting a vendor from a newer tech hub.

Kuwait Telecommunications Company (stc) announced a national initiative with the Kuwait Digital Academy to expand digital education and develop local talent, unveiled around the fifth Kuwait Digital Transformation Conference in early November. The programme aims to equip tens of thousands of people across Kuwait with essential AI skills in line with Kuwait Vision 2035. Designed to be inclusive, it starts with students in schools and universities and extends to professionals across the government and private sectors and to every segment of society, including senior citizens and people of determination. Through the partnership, stc and the academy will run workshops and training programmes covering programming, artificial intelligence and other modern technologies, with stc contributing its technical expertise to build engaging educational experiences. The initiative sits within stc's corporate social responsibility framework.

A lot of AI strategy is about chips and models; far less is about making sure ordinary people can use the technology. A telecom tying a large, deliberately inclusive AI-literacy drive to its national vision, and reaching groups usually left out like seniors and people of determination, is a quietly important way to make sure the benefits are widely shared.

A Kuwaiti university student or a mid-career government employee can enroll in a free workshop to learn practical AI and programming skills, gaining capabilities that make them more employable as public and private bodies across the country roll out AI tools.

OmniOps, a Saudi sovereign AI infrastructure company, announced the launch of Bunyan Inference Hub on November 11, a platform giving governments, enterprises, startups and developers access to open-source AI models and computing resources with data and workloads hosted in Saudi Arabia. The company noted that as organizations move AI into production, the need for different models can mean juggling separate providers, environments and integrations, and that bodies with strict data-residency or security requirements often felt the only option was to build and manage everything themselves. Bunyan offers a simplified platform they can plug into instead. The launch comes as Saudi-based entities have increased AI spending 124% year on year, per ServiceNow research cited by OmniOps, with Gartner forecasting global AI inference spending will reach $23.3 billion by the end of 2026.

Running AI in production is where data-residency and complexity problems bite hardest, especially for governments and regulated firms. A Saudi company offering a single, locally hosted gateway to open-source models and compute lowers the barrier for regional organizations to deploy AI without surrendering control of their data or stitching together foreign services.

A Saudi hospital or ministry that cannot send patient or citizen data to an overseas cloud can run open-source models through Bunyan, keeping both the compute and the sensitive data inside the Kingdom while still getting production-grade AI.

Abu Dhabi launched a new R&D Talent Enablement Program on November 11 to guide Emirati students from the classroom toward research careers in science, technology and engineering, part of a national push to build homegrown research talent. University students and recent graduates across the UAE can compete for prizes of 50,000 dirhams by developing working prototypes that solve real challenges, encouraging hands-on problem-solving rather than purely theoretical study. The programme is designed to shape the next generation of Emirati scientists and engineers and feeds a broader effort to strengthen the country's scientific research and advanced-technology base, as the UAE expands its global scientific presence and deepens investment in AI and deep-tech skills.

Sustained leadership in AI and advanced technology depends on a pipeline of local researchers, not just imported experts or purchased tools. A government programme that pulls students toward research careers and rewards building real prototypes is the unglamorous, long-term work that decides whether a country can keep innovating a decade from now.

A UAE engineering undergraduate with an idea for an AI tool can enter the programme, build a working prototype with guidance and funding, and use both the prize and the experience as a launchpad into a research career or a startup instead of leaving the field after graduation.

Omani fintech mamun has raised $3 million in seed funding led by Phaze Ventures through its Oman Fund, with participation from UAE-based Jasoor Ventures, reported in early November. Founded in 2019 and based in Muscat with entities in Abu Dhabi and Hong Kong, mamun operates a Shariah-compliant trade-finance platform connecting businesses with investors to finance imports, exports and working capital, and is licensed by Oman's Financial Services Authority under its crowdfunding framework. Its platform uses AI-powered underwriting to assess businesses, buyers, contracts, shipping documents and trading histories, tackling a global trade-finance gap the Asian Development Bank puts at $2.5 trillion, with 41% of small-business trade-finance applications turned down by banks. Around 40% of mamun's settlements already clear through Hong Kong institutions, and the funding will scale its underwriting along a Gulf-to-Asia corridor.

Trade finance is where small Gulf businesses most often hit a wall, rejected by banks that find them too costly to assess. An Omani company using AI to underwrite the very deals banks turn away, and focusing on the fast-growing Gulf-to-Asia corridor, shows regional AI being pointed at a concrete economic problem rather than a demo.

A small Omani exporter shipping goods to Asia that a traditional bank would reject can apply through mamun, whose AI assesses its contracts, buyers and shipping documents to arrange financing, giving the business the working capital to fulfill an order it otherwise could not fund.