On September 2, South Korea’s AI and software industry association announced that Aramco Digital, the digital arm of energy giant Saudi Aramco, had completed an industrial AI trial with seven Korean companies. The trial linked Korean AI models and chips with cloud infrastructure to test how engineering drawings and documents could be interpreted and turned into three-dimensional models. The announcement described the technology but left important commercial questions unanswered: Who would own products developed together and how would revenues earned abroad be shared?
For Gulf governments exploring such collaborations, starting with Saudi Arabia and the United Arab Emirates, those questions should guide the next round of AI partnerships with South Korea and Japan. Buying alternative technology can diversify suppliers. Building products together can also create engineering capabilities and export income. Gulf capital and industrial demand should secure a negotiated share in those products, with regional development teams and agreed rights to serve customers elsewhere.
The political opening already exists. The November 2025 Korea-UAE joint declaration envisages “joint investment, development, and export initiatives” in AI and semiconductors. A subsequent agreement between Korean firms and Aramco Digital has now produced a technical demonstration. These announcements describe cooperation but do not specify ownership of commercial products or overseas revenue rights. Those terms deserve attention before demonstrations become long-term supply contracts.
The Gulf also has leverage. South Korea’s science ministry has said it wants to make the consortium approach used with Aramco Digital a standard model for future overseas expansion. A successful Gulf deployment could help Korean firms win customers elsewhere, giving Gulf negotiators something to trade.
Agreements with South Korea offer the Gulf the clearest starting point. Japanese participation, including SoftBank’s place in Stargate UAE, the planned 1-gigawatt computing cluster led by Abu Dhabi’s G42 AI development holding company, creates another opening, although consortium membership alone says little about product ownership. South Korea and Japan also belong to Pax Silica, the U.S.-led technology supply-chain grouping that Qatar and the UAE joined in January 2026. Their shared participation provides a political basis for cooperation, while access to American chips remains subject to U.S. export controls. Three structures stand out.
Own the product together
The first is a Gulf-based product company owned by the partners. A Saudi industrial company and Korean AI firms could develop software that turns engineering drawings into usable digital models, drawing on the Aramco trial. The Saudi partner would bring operating know-how and controlled access to its data and would be the first customer. The Korean partners would contribute engineering and technology. A team based in the kingdom would adapt the resulting product and sell it.
The Gulf already has relevant precedents. In 2020, the Abu Dhabi National Oil Company (ADNOC) and G42 set up AIQ as a joint venture to develop and commercialize AI products for the oil and gas industry, combining ADNOC’s data with G42’s AI engineering. There is also an older cross-border precedent. The Korea Electric Power Corporation (KEPCO), which won the $20 billion prime contract for the Barakah nuclear plant in 2009, took 18 percent stakes in the plant’s commercial and operating companies in 2016.
The difficult negotiation concerns the boundary between existing technology and new work. Suppliers should retain their underlying models and tools, licensed to the venture on agreed terms. The venture could own the application developed with joint funding using those tools, with explicit rights to adapt it and sell it abroad. Its shareholders would also need to agree how costs and revenues are divided and what happens if one partner leaves. Sensitive industrial data could remain with its owner under defined access rules.
The commercial bargain must work for the engineering partner too. A Gulf customer could offer a paid development contract and access to a demanding industrial setting in exchange for shared rights in a reusable product. A successful demonstration would then open the door to other customers. Without that prospect, a supplier has little reason to accept the cost and complexity of a joint company.
There are obstacles on both sides. An East Asian supplier needs to protect technology used across many customers, while South Korean controls on designated national core technologies require approval or a filing before they are transferred abroad. A Gulf buyer needs enough engineering authority to avoid paying indefinitely for every modification. Negotiators should define the product and licensed technology early.
Tie investment to development
The second structure links Gulf equity investment in East Asian firms to a specific development mandate. Wa’ed Ventures—Aramco’s venture capital arm—funded a $15 million round for the Korean AI chipmaker Rebellions in 2024, which went on to establish a Saudi subsidiary in August 2025. Rebellions has described the Riyadh unit as a base for supplying Aramco’s data centers and for local sales and support. A development mandate would have to go further, for example, by requiring a funded product team in Riyadh or Abu Dhabi that is responsible for a defined application.
The equity stake and the development agreement would serve different purposes. Shares give the investor an interest in the company’s overall business, while a separate agreement would define the Gulf team’s responsibilities and any rights to the products it creates. Combining the two would let a fund support an existing firm without requiring a new joint venture for every investment.
In September, the chief executive of HUMAIN, Saudi Arabia’s state-backed AI company, said its planned venture fund would back companies that commit to running some of their computing in Saudi data centers or bringing staff to the kingdom. Those conditions could bring computing activity or staff to the kingdom; the reported investment criteria do not specify ownership rights in locally developed products. Gulf investors could go further in selected Korean and Japanese deals by agreeing what the regional team would develop and which commercial rights would accompany its work.
Investors seeking financial returns and governments seeking local capabilities may value the same commitment differently. Japan’s foreign investment screening framework also makes governance rights and access to sensitive technology important considerations, especially for state-backed investors.
Share the next contract
The third structure is a joint bid for government AI contracts in other countries, including in Africa and Central Asia. The UAE’s November 2025 commitment of $1 billion for AI development in Africa, implemented through the Abu Dhabi Exports Office with the UAE Foreign Aid Agency, offers one possible channel. Gulf and East Asian partners could combine financing and engineering with a recipient-country partner responsible for local delivery and support.
The consortium could also adapt and sell a product developed through one of the earlier structures. Its first opportunity should be a defined service with a credible budget, such as maintaining public infrastructure, rather than an entire national AI system.
Financing rules complicate the consortium. Japan’s STEP development-loan program, which supports the use of Japanese technology, permits certain joint ventures with recipient-country companies but requires Japanese leadership and a majority Japanese work share. The UAE initiative, meanwhile, aims to enable Emirati enterprises to deliver projects. Payment risk needs equal care: In May 2026, reports described G42 and Microsoft’s Kenya data center plan stalling over power constraints and a government unable to commit to leasing a set amount of computing capacity. Partners could explore co-financing a joint bid, with each lender financing work from its own country, subject to procurement rules and lender approval. They would still need to agree before bidding who bears delivery and payment risk.
For Abu Dhabi and Riyadh, the immediate task is to pick a few applications with real demand and settle these terms before scaling up. Each pilot agreement should state who develops the product and who may sell it abroad. Success should be measured by working teams, repeat customers and deployed infrastructure.
The September Aramco trial shows that technical cooperation is advancing. The next agreement should make its commercial consequences equally clear. Seoul wants the Gulf as its showroom, and the Gulf should set the price of that role in ownership.
