Government AI programs fail not because of flawed algorithms, but because agencies overlook the true cost of deployment, which accounts for 80% to 85% of total spending, according to a new analysis by Agile Dynamics released on 28 July 2026. The vendor quote typically covers only the visible layer - models, platforms, and cloud compute - leaving the bulk of the work unfunded until the program is already underway.
The real cost breakdown
Analysis of multiple government AI initiatives reveals four cost categories that repeat with enough consistency to plan against. AI models and platforms consume 15% to 20% of the budget. This is the part procurement teams estimate most comfortably, because it resembles software licensing. It is also the smallest share of any program that moves beyond a pilot.
Data preparation and engineering takes 25% to 30%. Government data rarely sits in a state ready for algorithmic use. It lives in formats designed for reporting, spread across ministries with different classification schemes, update cycles, and digitization levels. That data must be located, cleaned, structured, and piped through secure channels that respect privacy and classification rules - and then maintained continuously.
The largest share, at 35% to 45%, is integration and operational transition. This covers connecting AI output to the systems civil servants and citizens already use: API development, middleware, workflow redesign, and the period when old and new processes run side by side. A customs officer who has spent two decades assessing risk on instinct and documentary evidence cannot be expected to trust an algorithmic score without training, recourse procedures, and clarity on what to do when the score conflicts with their judgment. That is not resistance to change; it is a legitimate operational requirement that costs money.
Finally, governance architecture accounts for 10% to 15%. This includes audit mechanisms, bias testing, explainability frameworks, recourse procedures, and the monitoring that keeps the system performing as intended after deployment. In jurisdictions that have scaled AI successfully, this layer is designed before deployment, not bolted on afterward by the legal department.
Why procurement fails for AI
The standard procurement model treats the vendor quote as the baseline and everything else as margin or contingency. In AI programs, that quote covers only the first category. The remaining 80% to 85% is either invisible in the initial business case or scattered across budgets managed by different departments: data preparation assigned to IT as maintenance, operational transition filed under HR training with timelines that never align with the technology rollout, governance left to a legal team without the technical capacity to judge whether the system is auditable in practice.
The result is rarely outright failure. More often, the program is technically functional but operationally underfunded. The AI works. The integration is technically complete. But civil servants do not use it, or use it in ways that create more manual work than the process it replaced, or the system produces outputs that cannot be defended when challenged by a citizen or an auditor. The program then enters a cycle of supplementary funding requests, scope revisions, and delayed benefits - institutional fatigue that could have been avoided if the full cost structure had been visible from the start.
The dual-running architecture approach
The AI-native dual-running architecture method (ANDRA) starts from a thesis that applies with particular force in government: AI-native is a coexistence problem, not a replacement problem. No ministry moves from legacy to AI-native in a single cutover. Paper-era records, decades-old case management systems, and human-led decision processes keep running while AI capabilities come online. ANDRA treats the four cost categories as the price of running four operating states - legacy operations, AI overlays, redesigned value streams, and the AI-native target - in parallel under one governed architecture.
The mapping is direct. Data preparation becomes the executable company brain: turning policies, eligibility rules, and case histories from PDFs and spreadsheets into governed, queryable, machine-legible assets with named owners and audit trails. Integration and transition is the explicit design for the period when a caseworker's old workflow and the AI-assisted one run side by side, with every legacy system classified as retain, wrap, modernize, or retire. Governance is runtime governance: agent authority made explicit, audited, and reversible, with a named human principal and a kill switch for every system that acts.
Two real-world examples
In one diagnostic of a large regulated institution, the composite readiness score looked respectable until the layer breakdown showed institutional knowledge scoring lowest of all: policies in PDFs, pricing rules in spreadsheets, regulatory obligations living in the heads of experienced officers. Until that layer moved, no AI system could operate reliably in a regulated context - and every dirham spent on models ahead of it was spent out of sequence. The government parallel is exact: a permit ministry that licenses a language model before its regulations are machine-legible has bought the 15% and deferred the 30%.
The second example runs the other way. Federal HR authorities that completed structured process redesign before technology implementation have reported two to three times higher returns on subsequent system investments than peers that went technology-first. The jurisdictions most often cited for scaled government AI - Dubai, Singapore and several Nordic states - share the same structural feature: the full cost structure was budgeted and governed as a single program, not as a technology purchase with ancillary costs scattered across departments. The common element is not superior technology. It is architecture discipline.
What this means for the budget
For a CFO or procurement board reviewing an AI business case, the vendor quote should be treated as a component cost, not as the program baseline. The business case must account for all four categories, calibrated to the maturity of existing data infrastructure and the complexity of operational integration. A ministry that has already invested in unified data architecture will spend less on preparation and integration; one starting from fragmented legacy systems should expect those shares at the top of the range. Neither is better or worse - they are different starting points requiring different budget structures.
Vendor evaluation follows the same logic. A low license fee that demands extensive custom integration can cost more over the program lifecycle than a higher fee that ships with standard integration tooling and transition support. Total cost of ownership means all four categories, not the headline price. CFOs and procurement leads can build the necessary skills through resources like an AI Learning Path for CFOs to ensure business cases reflect the full cost picture.
When the four readiness assessments - technology, data, integration, governance - are completed before procurement begins, the budget that emerges is usually larger than the initial estimate, but it is also accurate. Programs built on that structure reach operational scale without the mid-program funding cycle that erodes political capital and institutional patience.
Why this matters for government finance and IT leaders
The cost of government AI is not in the algorithm. It is in the data, the integration, the governance, and the operational transition that surround it. Budgeting government AI as a dual-running architecture program - not a software purchase - prevents the most common failure mode: pilots that succeed technically yet never find the resources to become operational capabilities. Treat the vendor quote as one component among four, complete data, integration, and governance readiness assessments before procurement, and govern all four cost categories as a single program. That is the difference between a technically successful pilot and an operational capability. For a deeper look at the topic, see AI for Government.
Your membership also unlocks:
