From automating the Public Investment Fund to Dammam computing, the public investment pillars to turn artificial intelligence into a national industry

Artificial intelligence in the Saudi strategy has moved from a framework of supporting operational tools to, under Vision 2030, a fully integrated national industry that attracts domestic and international capital. With the declaration of 2026 as the Kingdom’s year of artificial intelligence, this step is based on documented achievements up to the end of 2025, where Saudi Arabia ranked first globally in digital governance and cybersecurity indices, and third globally in AI models. The path is led by the state’s investment arm, the Public Investment Fund, through a three-pillar plan that focuses on global investment to cement the Kingdom’s position as an innovation hub, integrate advanced technologies into its internal operations to become a data-driven institution, and accelerate the adoption of smart solutions across its entire portfolio of subsidiaries.

The establishment of Humain embodies the executive arm of this transformation through infrastructure and language models.The fund launched HUMAIN as a global entity specialized in building sector systems, deploying next-generation data centers, providing cloud-computing infrastructure, and delivering advanced models that protect local intellectual property while attracting talent and capital. At the core of this technical output is the ALLaM model, built entirely within the Kingdom by Saudi experts after being trained on more than 380 billion Arabic words, 394 thousand books and 500 billion data units. This intensive training enabled the model to grasp cultural contexts and regional dialects, earning it second place among Arabic models developed in the Arab world according to an independent evaluation by the Cohere platform on the Multi-Task Language Understanding (MMLU) benchmark. The HUMAIN Chat application relies on the ALLaM model to provide intelligent-assistant services targeting over 400 million Arabic speakers and two billion Muslims worldwide.

In terms of hardware and computing capacity, HUMAIN’s infrastructure relies on modern facilities and partnerships with international technology leaders, supported by a three-billion-dollar investment from Blackstone. Based in Dammam, the company manages infrastructure services for more than 150 countries, after building the world’s largest AI data center with a capacity of up to six gigawatts. The company operates under a unique model as the first global firm that does not purchase software from third parties, relying entirely on its self-developed system.

The fund’s internal operational impact provides a precise quantitative measure of the viability of agent-based automation.During 2025, the Public Investment Fund boosted its operational efficiency by deploying more than one hundred applications and activating 43 AI solutions in data analysis, task automation and decision-support. Through a unified, secure corporate workspace, the fund integrated 15 AI agents serving over 2,300 users. By the end of 2025 this effort had automated more than 1,500 manual procedures and workflows across the fund’s system, contributing to a saving of over three million work hours.

This transformation reshapes the practical reality for technical teams and decision-makers across the Gulf, Egypt and the Arab Levant. Operating a six-gigawatt data center in Dammam alters the equation for hosting large models for regional companies, reducing mandatory reliance on distant Western cloud paths and giving financial and governmental institutions flexibility that aligns with data-sovereignty standards and low latency. Moreover, the entry of a competitive Arabic model built on hundreds of billions of Arabic words forces product managers in Riyadh, Cairo and Amman to reconsider inference budgets, shifting from full reliance on superficially adapted foreign model APIs to solutions that understand linguistic and regulatory privacy. The model delivered by the fund, providing three million hours through 15 intelligent agents, moves evaluation standards in regional companies from formal AI pilots to strict productivity calculations and reduced institutional waste.