AI Services Case Study — Developing an AI Business Case for Scaling Generative AI Licences
- Jan 23
- 7 min read
Updated: 3 days ago
1. Organisational Problem
Following a successful pilot of generative AI tools within selected teams, a UK public sector organisation began exploring the purchase of licences to enable wider use across the workforce.
Initial feedback from the pilot was positive. Staff reported productivity improvements in drafting, summarisation and research activities, and there was growing enthusiasm among senior leaders to scale adoption more broadly.
An initial business case for generative AI licences was developed and approved in principle, based on relatively low unit costs and perceived productivity benefits.
However, as the proposal progressed, the organisation’s Audit and Risk Committee raised a number of concerns regarding the robustness of the business case and the absence of supporting governance and strategic context.
In particular, the Committee noted that the proposal had been developed without:
A defined AI Strategy or Roadmap
An understanding of organisational AI Capability and Maturity
Structured AI Use Case Discovery to identify priority applications
Appropriate governance and assurance arrangements
While the pilot had demonstrated potential, there was no overarching structure to guide how generative AI should be adopted, governed or scaled across the organisation.
In the Orr Consulting AI Transformation Process, this case study demonstrates the Design-stage role of AI Business Case development in testing whether proposed AI investment is strategically aligned, affordable, governable, deliverable and capable of producing measurable value.
2. Situation
The organisation operated within a complex public sector environment, with a strong focus on accountability, value for money and risk management.
Following the pilot, there was increasing momentum to scale access to generative AI tools across a wider group of users.
The organisation had already completed a structured pilot process in which benefits were defined, tracked and evidenced. This provided a more credible evidence base for the business case and helped shift the discussion from initial enthusiasm to informed consideration of wider rollout.
While the cost per licence appeared relatively low, the potential scale of deployment meant that total investment would be material. In addition, the organisation would need to demonstrate that any investment represented appropriate use of public funds.
The organisation also faced a common public sector constraint: while one-off capital funding could potentially support initial deployment, committing to ongoing recurring licence costs required a higher level of scrutiny and long-term justification.
The Audit and Risk Committee concluded that the proposal did not yet meet the standard expected for a material organisational investment.
In particular, the Committee highlighted:
No defined AI Strategy or Roadmap
No AI Capability and Maturity baseline
No formal governance framework or acceptable use policy
Limited assessment of data protection, information security and Shadow AI risks
The Committee therefore concluded that the proposal treated AI as a procurement decision without sufficient consideration of the broader organisational implications of adoption.
The leadership question therefore became:
"How do we move from a successful pilot to controlled, value-driven and governable adoption of generative AI?"
3. Background
Like many public sector organisations, the authority was experiencing growing exposure to generative AI through both formal pilots and informal experimentation.
The initial business case focused primarily on licence procurement and assumed that benefits would emerge through increased access to AI tools. However, the Audit and Risk Committee identified that the underlying issue was not the procurement of licences, but the absence of a structured approach to AI adoption.
A more comprehensive business case was therefore required to test value, risk, governance, capability and organisational readiness before wider investment could be considered.
The organisation therefore engaged Orr Consulting to facilitate the development of a revised AI Business Case capable of supporting a more informed and defensible investment decision.
4. Action Taken
Orr Consulting was engaged to facilitate and advise on the development of a structured AI Business Case, ensuring that the organisation retained ownership of both the outputs and the decisions required for delivery.
The engagement reframed the proposal from a straightforward licence purchase into a broader investment decision concerning how Generative AI should be introduced, governed, funded, adopted and scaled across the organisation.
Using Orr Consulting’s AI Business Case Development methodology, the proposal was reassessed across the following areas.
4.1 Deliverables and Timeline
The business case defined a targeted initial rollout of generative AI licences to priority user groups, supported by governance controls, training and clear milestones for deployment, review and benefits monitoring.
4.2 Cost and Investment Profile
A fuller view of cost was developed, covering licence expenditure, training and enablement, governance implementation and ongoing support.
This showed that, although unit costs appeared low, the total cost of ownership at organisational scale was more significant. Particular attention was given to the recurring nature of licence costs, which required clear justification in the context of public sector funding constraints and long-term affordability.
This prevented the affordability assessment from being based solely on an attractive per-user licence price while overlooking the cumulative and continuing organisational commitment.
4.3 Benefits Realisation
Expected benefits were defined in relation to prioritised use cases, including productivity improvement, greater consistency of outputs, faster turnaround of communications and better reuse of organisational knowledge.
The business case also considered what benefits were expected, where they would arise, when they would be realised, who would be accountable and how they would be measured. In the public sector context, ongoing recurring licence costs would need to be supported by evidenced efficiency savings and demonstrable operational benefits over time.
This provided a basis for defining, measuring and tracking benefits through a structured AI Benefits Realisation approach as deployment progressed.
4.4 Risk Identification
A structured risk assessment considered data exposure, inappropriate reliance on AI-generated outputs, inconsistent usage, unclear accountability and the continued emergence of Shadow AI.
To mitigate these risks, the business case incorporated the introduction of the Orr Consulting AI Governance and Assurance Framework, including an Acceptable Use Policy, defined oversight and accountability, structured risk processes and targeted staff education.
4.5 Alignment with Strategy
The business case was explicitly aligned to organisational priorities relating to service improvement, efficiency and value for money, and to the emerging direction of the organisation’s AI strategy.
This avoided the risk of treating licence procurement as an isolated technology decision and instead positioned it as part of a more structured approach to AI adoption.
4.6 Delivery Capability
The organisation’s AI capability and maturity were considered to assess its readiness to adopt generative AI at scale.
This highlighted gaps in governance, organisational understanding and delivery capability, confirming that the investment decision needed to reflect not only the potential value of the technology, but the organisation’s ability to implement and sustain it in a controlled way.
4.7 Options Appraisal
The business case evaluated three broad options:
Do nothing and continue informal use of external tools
Limited targeted rollout to specific user groups
Broader organisational rollout with governance controls
The options appraisal ensured that the business case did not treat broader rollout as the default or inevitable outcome of a successful pilot.
This enabled leadership to weigh value, risk and pace of adoption before proceeding.
5. Outcomes
The business case engagement created several notable outcomes, observations and lessons learned.
5.1 Refined Investment Decision
Rather than proceeding with a broad rollout, the organisation agreed a targeted initial deployment of generative AI licences focused on priority user groups aligned to high-value use cases.
This allowed the organisation to preserve the value demonstrated through the pilot while limiting initial exposure and creating a stronger evidence base for any later expansion.
5.2 Governance Established
The revised business case confirmed that deployment of Generative AI licences would be accompanied by implementation of a formal AI Governance and Assurance Framework.
This ensured that access to approved tools would be supported by clear policies, accountability, oversight and risk-management arrangements.
5.3 Strategic Alignment
Structured AI Use Case Discovery connected the proposed investment to specific organisational objectives, priority user groups and credible operational applications.
This reduced the risk of purchasing licences broadly and then expecting staff to identify value after deployment.
5.4 Value for Money
The organisation was able to demonstrate a clearer link between investment, expected benefits and organisational outcomes, supporting public sector accountability requirements.
The revised case also made clear that value for money would depend on evidenced adoption and benefits over time, rather than on licence deployment alone.
5.5 Reduction of Shadow AI Risk
By providing approved tools alongside clear governance, education and guidance, the organisation reduced reliance on uncontrolled external AI tools.
This demonstrated that Shadow AI risk could be addressed through a proportionate combination of enablement and control rather than through prohibition alone.
5.6 Governance Gateway
The engagement demonstrated that an AI Business Case should do more than document cost and expected benefits.
It provided a formal gateway through which strategic alignment, affordability, governance, capability, risk, delivery readiness and expected value could be tested before wider organisational commitment.
Without this gateway, the organisation risked allowing pilot enthusiasm and apparently low unit costs to drive a material recurring investment without sufficient evidence or control.
5.7 Implementation Path
Following approval of the revised business case, the organisation agreed a controlled implementation path comprising:
implementation of the AI Governance and Assurance Framework alongside licence deployment;
targeted training and awareness for priority users;
monitoring of usage and benefits associated with prioritised use cases;
refinement and expansion of the use-case portfolio based on evidence;
use of deployment learning to inform broader AI Strategy and Roadmap development.
This created a proportionate route from successful pilot evidence to controlled deployment, review and potential future scale.
6. Final Thoughts
Generative AI licences are often perceived as a low-cost and relatively straightforward entry point into AI adoption.
However, organisational deployment is not simply a procurement decision. It creates recurring financial commitments and introduces wider questions of strategic alignment, governance, capability, adoption, risk and measurable value.
In this case, the structured AI Business Case acted as a critical governance gateway between successful pilot activity and wider delivery. It required the organisation to test the proposed investment more fully before committing to broader rollout.
The resulting decision was not to abandon the opportunity or proceed immediately at scale. It was to approve a targeted and governed deployment focused on priority users and evidenced use cases.
The most important outcome was not the approval of licences. It was the organisation’s recognition that successful AI adoption requires a structured investment and transformation decision rather than incremental procurement.
This AI Services Case Study is part of the Orr Consulting AI Insights Library — structured thinking for AI transformation leaders and decision makers.
If your organisation would benefit from a clearer and more evidence-based assessment of a proposed AI investment, we would be pleased to discuss your next AI steps.
Subscribe to Orr Consulting to receive occasional emails with practical AI Insights and updates.
