Nowhere a vendor can look it up. Vision 2030 publishes targets, not budgets: the word budget appears once in its 2025 annual report, which runs to 297 printed pages. The Saudi state budget has nine sectors and none of them is technology. The only official government technology figure is published by the Digital Government Authority, and it is falling.

That is the honest answer to a question almost every European software company asks before its first trip to Riyadh. The rest of this article is about what is published instead, where to find it, and the two gates that decide whether a funded project ever reaches a contract.

All figures here are current as at 26 October 2026.

Vision 2030 does not have a budget you can look up

The Vision 2030 Annual Report 2025 is a serious document. It is also not a financial document. Across its 297 printed pages the word "budget" appears exactly once, and it appears as a strategic objective of the National Transport and Logistics Strategy: "Contribute to achieving a balanced public budget." In the 92 page National Transformation Program Annual Report 2025 the word does not appear at all.

What the report publishes instead is a measurement architecture, and it says so: "Performance indicators are defined across three levels. The first and second levels measure progress toward strategic objectives, while the third level tracks the implementation of initiatives under Vision Realization Programs and national strategies." Three levels of indicators. None of money.

The Vision Realization Programmes are described as "primary implementation tools" and "medium-term entities" that have "strengthened coordination across the ecosystem." The National Transformation Program oversees more than 313 initiatives, leads 34 strategic objectives, measures 78 strategic indicators and "collaborates with 7 leading ministries and 50+ entities." Its committee is chaired by a member of the Council of Economic and Development Affairs and its members are sitting ministers, including the Minister of Communications and Information Technology and the Assistant Minister of Finance. That is a coordination forum of the people who hold budgets. It is not itself a budget holder.

So a vendor who builds a pipeline around "the Vision 2030 technology budget" is chasing an object that does not exist as a published number. The programme sets the direction and the targets. The money is appropriated somewhere else, by someone else, and it is reported somewhere else again.

One carve-out matters, because it is where the headline money actually is. The Public Investment Fund Program is itself a Vision Realization Programme, and the PIF does deploy capital directly: the report puts its assets under management at 909 billion US dollars and describes its AI vehicle HUMAIN. The PIF is a real counterparty that signs real contracts. It is also, as the next sections show, outside the government procurement system entirely.

The state budget has nine sectors, and technology is not one of them

Saudi Arabia allocates expenditure across nine sectors: Public Administration, Military, Security and Regional Administration, Municipal Services, Education, Health and Social Development, Economic Resources, Infrastructure and Transportation, and General Items. The list is closed. It sums to the total. There is no technology sector, no ICT sector and no digital sector.

Technology is not absent from the budget document, though, and this is the part most commentary gets wrong. It is named and then never costed. The FY2026 Budget Statement publishes the remit of each sector, and the Infrastructure and Transportation sector's remit expressly includes telecommunications and information technology, data and artificial intelligence, and digital government, alongside roads, transport safety, ports, airports, real estate, postal services, space and industrial cities. Cybersecurity sits in a different sector again, Security and Regional Administration.

Infrastructure and Transportation was allocated SAR 35 billion for 2026. It is the smallest of the nine sectors, and it carries all of the above. Meanwhile a line ministry's own systems spending is classified under that ministry's sector, so hospital software lands in Health and Social Development and school software in Education. There is no way to reconstruct a technology total from the published document, and the Ministry of Finance publishes no data against the communications category in its functional classification standards.

Two further things a vendor should know about the published budget.

The third largest of the nine sectors is General Items, at SAR 236 billion budgeted for 2026 and revised up to SAR 265 billion at mid year. The Ministry of Finance calls it, in its own parenthesis, the central spending sector. It carries pensions and social insurance, contributions to international organisations, government programmes and facilities, debt cost, the budget support reserve, the essential services support reserve and the emergency reserve, and subsidies to government entities. The categories are listed. No amount is attached to any of them. It is the largest opaque block in the budget, and the explanations for its growth that do get published name the giga projects: the Third Saudi Expansion of the Grand Mosque, and Qiddiya.

And for 2027 there is currently no sector breakdown at all. The FY2027 Pre-Budget Statement, published on 30 September 2026, estimates expenditure at about SAR 1,392 billion and revenue at about SAR 1,202 billion, with a deficit of about SAR 191 billion or 3.6 percent of GDP. It contains no sector table. Pre-budget statements never do. The sector figures arrive with the full Budget Statement, and the Ministry's own published timetable puts that in December: "final approval in October and November, and the official budget announcement takes place in December."

The fiscal context is worth stating plainly rather than reading into. Deficits are projected for 2026, 2027, 2028 and 2029, after actual deficits in 2024 and 2025. The 2026 deficit was revised from SAR 165 billion to around SAR 245 billion. But the documents do not describe retrenchment. There is no mention of deferral, postponement, reprioritisation or austerity anywhere in them. The Citizen's Version of the 2026 budget calls the policy expansionary and says the planned deficit is a strategic choice to finance productive projects. Capital expenditure rose 32.4 percent year on year in the first half of 2026. Our piece on which GCC country to enter first sets out the wider macro picture; the budget documents themselves are not an argument for a slower market.

The one official number: SAR 31.9 billion, and it is falling

There is exactly one published, official figure for Saudi government spending on technology, and it does not come from the Ministry of Finance. The Digital Government Authority publishes an annual Government Spending Report on ICT, prepared with the Ministry of Finance, the Expenditure and Projects Efficiency Authority, Monsha'at and the National Center for Government Resource Systems.

The 2025 edition, announced on 15 June 2026, reports that "total government spending reached SAR31.90 billion in 2025, while the value of government contracts amounted to approximately SAR31.70 billion across more than 6,145 contracts." SAR 31.9 billion is about 2.4 percent of total government expenditure for the year.

The same report gives the composition of the growth. Cloud computing spending rose 42 percent against 2024 and spending on artificial intelligence and emerging technologies rose 20 percent. Local content "accounted for 49% of government software procurement." Purchase orders through national framework agreements exceeded SAR 5.16 billion, "benefiting more than 500 government agencies and 65 companies." Small and medium enterprises reached 23 percent of spending, and contracts awarded to them came to approximately SAR 9.23 billion, 29 percent of the total value of government contracts in the sector.

Now the part that nobody repeats. On the Authority's own comparable measure, government ICT spending peaked in 2023 and has fallen every year since: SAR 34.84 billion in 2022, SAR 41.87 billion in 2023, SAR 37 billion in 2024, SAR 31.90 billion in 2025. That is a fall of roughly 24 percent from the peak, during the period when every forecast for the Saudi technology market was rising.

The Authority explains it, and the explanation is not a warning. The 2024 report attributes that year's decline "primarily due to the settlement of non-recurring financial dues related to services provided in previous years, most notably internet and digital circuit services." The 2025 report attributes its own: "This reflects the maturity of the digital government ecosystem and increased spending efficiency following the completion of several core digital infrastructure projects that represented a major component of government technology investments in previous years."

Read that as a buyer, not as an analyst. The money that has come out of the series is infrastructure and connectivity, the one off build. What is growing inside a shrinking total is cloud and AI. A vendor selling network capacity or data centre build into Saudi government is selling into a line that is closing. A vendor selling software, cloud services or applied AI is selling into the part that is still expanding, and competing for a share of a total that is smaller than the market forecasts imply.

One trap to avoid while you are in this material. The large technology numbers that appear in Saudi official publications are usually market size, not government budget: the whole-economy communications and technology market is an order of magnitude larger than government ICT spending, and the two get quoted interchangeably. Our piece on where GCC technology investment is going handles the market side. This section is the government side, and they are not the same pool.

What is published is the plan, not the price

Here is the single most useful published thing in the Saudi system, and almost no foreign vendor uses it.

Article 12 of the Government Tenders and Procurement Law requires every government agency to plan its purchases in advance and to publish that plan. In the Bureau of Experts official English translation: "A government agency shall plan in advance its works and procurements, and coordinate with the Ministry for the allocation of necessary appropriations. At the beginning of each fiscal year, the agency shall publish a plan consistent with its budget which includes key information about its works and procurements for the year, without compromising the confidentiality considerations of national security."

The Law says "at the beginning of each fiscal year." The deadline is tighter than that, but it is set by a lower instrument: Article 3 of the Executive Regulations, issued by Ministerial Decision No. 1242, fixes publication within the first quarter of each financial year. There is no official English text of the Executive Regulations in circulation, so the following is a translation of the Arabic rather than a quotation of the Ministry's own English. Article 3 requires the plan to carry, as a minimum, the type and nature of the works and procurements, the place of execution, and the method of tendering and contracting. National security items, weapons and military equipment are exempt. Publication is on the Portal and on the government agency's own website, and continues until the works are tendered. Plan information is to be updated continuously.

The Portal is Etimad, and the plans are genuinely public. Etimad's future tender projects page requires no login and no Saudi identity, and as at today it holds 95,725 published plan records. Each one names the agency, the project, its nature and description, the quarter and year it is expected in, its expected duration, whether it is inside or outside the Kingdom, and its project type, where information technology is a classified type in its own right.

Three honest limits, and the third is the interesting one.

The list is not usable as a pipeline tool out of the box. It is sorted oldest first, so the first page a visitor sees is 2020. There is no public filter by year and no public filter by information technology activity, and the API ignores both if you supply them, though filtering by agency works. The information is in the records. It is simply not queryable, which means the work of turning 95,725 rows into a named list of agencies planning software purchases in 2027 is manual, and it is therefore work almost nobody does.

Publication creates no commitment. Article 12(2): "The government agency's publication of its plans of works and procurements shall not result in any obligation." A published plan is an intention, not a tender.

And the plan contains no money. Not because of an oversight, and this is the point: the same Law makes the price confidential. Article 23 requires an agency, before estimating the cost of works and procurements, to "thoroughly study market prices and take them into consideration when setting a maximum limit for the expected total contract value" and to "set measures necessary to maintain the confidentiality of the estimated cost." So Saudi law obliges the buyer to publish what it intends to buy and obliges it to keep secret what it expects to pay. Every public forward plan in the Kingdom is a list of requirements with the budget deliberately removed. If you have been waiting for a published technology budget to plan against, that is the answer: the number exists, it is calculated, it is reviewed internally, and it is withheld by design.

Two agencies are outside this system altogether, and one of them matters enormously. The Law applies to all government agencies, with exemptions granted under Article 95, and Royal Decree No. M/76 expressly continues "the exemptions issued to each of the Public Investment Fund and the Saudi Central Bank." The PIF, which holds the headline technology and AI money, does not tender through Etimad and publishes no Article 12 plan.

Note the date on all of this. The Law quoted above is Royal Decree No. M/128 of 2019, which is still the operative law today, and it is being replaced on 2 January 2027. Two of the published amendments land directly on Article 12: agencies will have to coordinate with the Local Content and Government Procurement Authority during advance planning, and agencies with no line items in the State General Budget will no longer coordinate appropriations with the Ministry of Finance. The publication duty and the no obligation clause survive. We set out what else changes in our piece on Saudi market entry for technology companies.

The gate that is not in the procurement law

Suppose everything has gone right. An agency has an appropriation, it published a plan, it ran a tender, and it wants to buy your software. There is one more approval, it is specific to technology, and it is not in the procurement law at all.

Since a circular numbered 01/42/5589 and dated 07/11/1442H, a Saudi government entity has needed the Digital Government Authority's prior approval before establishing a new digital platform. The Authority's own announcement puts the scope beyond doubt: entities must "obtain prior approval from the Authority before establishing any new platform," and the platforms "include mobile applications and websites in all domains, integration channels, and digital services, whether they will be launched directly through the entity or a contracting company."

That requirement is now operated through the Authority's Controls for the Registration and Management of Digital Government Domains, Platforms, and Services, document DGA-1-2-2-106. The timing is in its definitions: a registration certificate is "granted before commencing the development of a new digital platform." Not before launch. Before development.

What makes this a commercial risk rather than paperwork is that the Authority can say no, for reasons that have nothing to do with the vendor or the price. Control 2-106-07 requires an entity to refrain from requesting a new platform that provides services "already available on other platforms, whether the requesting entity owns such platforms or not." Control 2-106-14 gives the Authority the right to reject a registration request where "the DGA has a strategic plan that conflicts with registering the platform, such as plans to merge digital platforms with similar scope and services," or where the entity has unresolved regulatory violations on its other platforms. Approval is also required before launching new digital services or modifying existing ones, and before decommissioning or modifying a platform.

The enforcement is not a fine. There is no penalty provision in the Authority's statute and none in the Controls, whose entire compliance clause is that the Authority "shall evaluate the extent of government entities' compliance with these controls and measure it." The Authority's own regulatory framework places its Controls below laws and policies on its enforcement hierarchy. Be precise about this if you are briefing a board: these are mandatory administrative instruments resting on a Council of Ministers resolution, not a law with sanctions.

The teeth are elsewhere, and they are sharp. No registration certificate, no domain: an entity "must obtain a digital platform registration certificate before reserving the domain name." And the Authority can suspend a non-compliant domain after 10 working days to rectify, delete it and revoke the entity's ownership after a further 10, and suspend a main domain even where the offending platform is an affiliated sub-platform.

So the Saudi duplication test is enforced by a third party that is not your buyer, at the design stage, on the question of whether government already has something like what you are selling. A tender win is not the last gate.

A large share of the budget sits inside government-owned companies

The last place to look is the one that does not appear in any budget line or procurement portal, because the buyer has been corporatised.

Elm Company was established on 9 June 1988 as a limited liability company with capital of SAR 500,000, converted on 13 November 2007 into a one-person closed joint stock company wholly owned by the Public Investment Fund under Royal Decree No. M/90, and listed on the Saudi Exchange in February 2022. It reported revenue of SAR 9,464,884,988 for 2025, up from SAR 7.41 billion in 2024, with business process outsourcing revenue up 43.3 percent. It builds and operates named national platforms including Absher, Muqeem, Yakeen, Tamm and Smart Gate, and in January 2025 it signed an agreement to acquire the Public Investment Fund's entire stake in Thiqah Business Services in a deal valued at SAR 3.4 billion, completion being subject to regulatory approvals. Thiqah runs its own vendors portal and its own qualification certificate, separate from Etimad.

Alongside them sits Tahakom, established in 2015 under a PIF subsidiary, which operates national camera and traffic systems and its own data processing centres, and the National Information Center, which is not a company but the operational arm of the Saudi Data and Artificial Intelligence Authority. The Center operates Nafath, the national identity layer every Saudi Etimad login passes through, and Deem Cloud, the government cloud platform. Its National Data Bank hosts more than forty government entities in its data lake with data integrated from over 240 government systems.

For a vendor, the practical consequence is that a material part of what looks like government technology demand is intermediated by companies that run their own supplier onboarding, buy on commercial terms, and are not bound to Etimad for their own purchasing. Those are addressable counterparties, and they are reachable without waiting for a tender. They are also, in many cases, your competitor for the same end customer.

What this means in practice

Six recommendations, in the order we would act on them.

Stop looking for a technology budget and start reading forward plans. The number you want does not exist and is withheld by law. The thing that does exist, and that your competitors are not reading, is 95,725 published plan records naming agencies, projects, quarters and project types. Build the list by agency, because agency filtering works, pick the twenty agencies that plausibly buy what you sell, and read their plans. Do it in the first quarter, which is when the Regulations require them to appear.

Sell into cloud, software and applied AI, and do not build a Saudi plan on infrastructure. The official series says government ICT spending has fallen about 24 percent from its 2023 peak while cloud rose 42 percent and AI rose 20 percent in a single year. That is a composition shift, not a slowdown, and it is published by the buyer.

Put the Digital Government Authority registration step in your own deal plan, before the tender. Ask the buying entity, early, whether the platform is registered and whether a registration certificate exists. Ask specifically whether any other government entity already offers the same service, because that is the ground on which the Authority refuses. A deal that fails this test fails after you have spent the pursuit cost.

Treat the Public Investment Fund and its portfolio companies as a separate motion. They are exempt from the procurement law, they publish no forward plan, and nothing you learn about Etimad helps you there. Elm, Thiqah and Tahakom are approached as commercial accounts through their own vendor portals, not as tenders.

Remember that a willing buyer is not sufficient on the foreign vendor question either. Contracting a party licensed outside the Kingdom for work inside it brings its own conditions, and separately the regional headquarters controls bear on government contracting; we cover the first in our Saudi market entry piece and the second in our explainer on the RHQ requirement. The forward plan tells you what is coming. It does not tell you that you are eligible.

Finally, be clear eyed about Saudi programme money. The National Technology Development Program runs a long list of funding products, and almost every one requires a Saudi commercial registration, a Saudi headquarters or Saudi nationality. The single product open to a company without one is Relocate, which offers grants up to SAR 7.5 million, is explicitly "for companies without existing headquarters or business activities in KSA," and requires you to relocate senior management to Saudi Arabia. That is not grant funding for an exporter. It is an incentive to stop being one, and it should be evaluated as a market entry decision rather than a subsidy.

If you are working out which agencies to approach and what your Saudi entry actually needs to look like, get in touch.

Sources

This is commercial guidance, not legal or tax advice. Saudi procurement, licensing and digital government rules change frequently, and the figures and provisions here are current as at 26 October 2026, with the Government Tenders and Procurement Law due to be replaced on 2 January 2027. Verify the position with qualified Saudi counsel before making a commitment.