Saudi AI tool use doubled in a year, and the regulator has published a guide for companies catching up

The Communications, Space and Technology Commission released an awareness guide for artificial intelligence adoption at technology companies on 3 August 2026, aligned with the Council of Ministers' designation of 2026 as the Year of AI. It is addressed to executive leaders, strategy directors and product managers, and is built on two pillars: AI applied inside internal business processes, and AI embedded into the products and services a company sells. The guide sets out a readiness framework across five dimensions — context, data, infrastructure, skills and expertise, and organisational culture. Separately, CST research put AI tool adoption among Saudi internet users at 45.2%, more than double the level recorded the previous year. The two-pillar split is the useful part, because those are genuinely different undertakings with different economics. Using AI internally is an efficiency programme: it changes cost, it can be piloted in one department, and its return shows up in operating expense. Embedding AI into a product is a product-development programme: it changes what the company sells, it requires the model to work for customers rather than colleagues, and its return shows up in revenue. Companies routinely conflate the two, and a firm that has automated some internal reporting will describe itself as an AI company. The guide's structure makes the distinction explicit before a strategy is written, which is the point at which it is cheapest to be clear about it. The five readiness dimensions do similar work. Three of them — data, infrastructure, and skills and expertise — are the ones organisations expect. The other two are the ones that tend to determine outcomes. Context asks whether the use case is one where AI offers a genuine advantage, which is the question that prevents a deployment nobody needed. Organisational culture asks whether the people whose work changes will actually adopt it, which is where most enterprise technology programmes are decided regardless of sector. The 45.2% adoption figure gives the guidance its commercial backdrop, and it should be read for what it measures. It covers individual internet users reaching for AI tools, not organisations deploying them. Doubling in a year is a rapid consumer-side shift, and it is consistent with the finding cited at DeepFest this week from Stanford's AI Index, that more than 80% of surveyed employees in Saudi Arabia use AI regularly in daily work. Both figures describe the same underlying condition: individual familiarity is high and rising quickly, while the institutional layer is where the work now sits. That is a favourable starting position, and an unusual one. Most markets have to build individual comfort with these tools before organisations can deploy them. Saudi organisations are starting with a workforce that already uses them. The guidance also sits inside a fuller regulatory architecture than the technology-company audience might suggest. SDAIA's AI Adoption Framework, released in November 2025, is a mandatory baseline for public sector entities across five governance pillars, and reaches private suppliers through procurement rather than statute. A draft Responsible AI Policy went to public consultation with submissions due by 3 May 2026. The Personal Data Protection Law is fully enforced and carries penalties of up to SAR 5m per breach. Against that, CST's contribution is advisory rather than binding, and aimed at a different audience. SDAIA's framework tells a ministry what it must do before deploying a model. CST's guide tells a technology company how to think about building one. A supplier that reads both is looking at the same requirement from the buyer's side and its own, which is the practical reason to read them together. The guide predates LEAP by four weeks, so it is not an event announcement. It is the document technology companies were given before arriving at a show where a great deal of AI product was launched, and that sequencing is deliberate enough to be worth noting.