Establishing a Generative AI Implementation PMO for One of the Largest Banks

The Bank is one of the largest banks in its market, launching a Generative AI programme spanning multiple use cases across customer service, operations, and internal productivity. Given the scale, novelty, and risk profile of generative AI initiatives, the Bank required a dedicated programme management office to govern implementation.

• Multiple generative AI use cases being explored concurrently without a coordinated programme structure
• Limited standardised approach to evaluating, prioritising, and sequencing generative AI use cases
• Novel risk considerations specific to generative AI, such as model governance and responsible AI use, not yet formally addressed
• Fragmented ownership across business and technology teams for generative AI initiative delivery
• No consistent framework for tracking value realisation and adoption of generative AI use cases

• Generative AI adoption in banking is moving quickly, and disciplined programme governance is essential to capture value while managing model and regulatory risk responsibly
• Without coordinated prioritisation, generative AI investment risks being spread across too many use cases with insufficient depth to realise value in any one of them
• Responsible AI governance is increasingly a regulatory and reputational expectation for financial institutions deploying generative AI at scale
• Left without a dedicated PMO, generative AI initiatives tend to stall in pilot phase, with no clear path to scaled, value-generating deployment

• Generative AI Programme Management Office Structure
• Use Case Prioritisation & Sequencing Framework
• Responsible AI Governance Framework
• Model Risk & Data Privacy Oversight Process
• Value Realisation Tracking Framework
• Programme Status Reporting Dashboard
• Phased Rollout Roadmap

100%

Generative AI use cases brought under unified programme governance

1

Responsible AI governance framework established (approx.)

20–30%

Improvement in use case prioritisation and resourcing efficiency (approx.)

90%+

Use cases assessed against the new prioritisation framework (approx.)

25%

Faster progression of prioritised use cases from pilot to scaled deployment (approx.)

1

Value realisation tracking framework deployed programme-wide (approx.)

The Bank now has a governance foundation for pursuing generative AI opportunities responsibly and at scale, beyond programme oversight for the initial rollout. Structured prioritisation, responsible AI governance, and value tracking let the Bank keep expanding its generative AI portfolio with discipline.