From strategy to scale: Building data and AI capability for the project delivery profession

Over the past three years, the project delivery profession has undertaken an ambitious journey to embed data, artificial intelligence and innovation into the heart of government project delivery.

What began with a vision set out in the Data analytics and AI Strategy has evolved into a programme that has helped lay the foundations for a more data-driven, AI-enabled future across the profession. Today, as the formal Data and Artificial Intelligence (AI) Programme reaches its conclusion, there is an opportunity not just to reflect on what has been achieved, but to focus on the next challenge: turning new capabilities into everyday practice.

The journey began in 2024 with the publication of ‘Data Analytics and AI in Government Project Delivery‘, fondly known as the Belfast Paper, the first function-specific strategy focused on data and AI. Alongside this, a cross-function steering group was established, bringing together expertise from across government and industry to help shape the agenda. From the outset, the ambition was clear: to ensure the project delivery profession could harness emerging technologies and better-quality data to improve delivery outcomes.

Momentum accelerated over 2024 and into 2025 as the wider government landscape embraced the opportunities presented by artificial intelligence, the profession found itself in a strong position. New digital and data competencies were incorporated into the Project Delivery Competency Framework, helping ensure that data literacy and digital skills became an integral part of professional development.
At the same time, attention turned to practical experimentation. More than 20 AI pilots were identified across the system and the first “Show and Tell” events created a space for practitioners to share insights, challenges and lessons learned. These initiatives reflected a core belief that innovation succeeds when people are encouraged to learn from each other and build on what works.

One of the programme’s most significant achievements has been the development of Assure Assist (previously known as Scout), an AI-enabled tool designed to support assurance activity. By June 2025, Assure Assist had successfully completed beta testing and demonstrated an accuracy rate of 92.5 per cent in identifying appropriate lines of enquiry. The promise of the tool was recognised shortly afterwards when the programme secured funding from the Government Efficiencies Accelerator Fund to support its continued development and scaling. Assure Assist is now secure by design, stable, and ready to support Government Major Projects Portfolio assurance reviews.

However, technology alone does not drive transformation; people do. Throughout the programme, a strong emphasis was placed on building confidence and capability across the profession. Project delivery-specific learning offers were introduced, including the project delivery “AI for All” offer and the launch of a dedicated prompt library hosted on the i.AI Knowledge Hub. Alongside these initiatives, the profession hosted a series of Decoding Delivery webinars, attracting thousands of viewers and helping to demystify AI experimentation for project professionals across government.

The programme also invested heavily in leadership capability. Dedicated Major Projects Leadership Academy digital, data and AI masterclasses and alumni workshops have now been delivered to 250 senior major projects leaders, the Project Leadership Programme has delivered data and AI workshops to over 180 alumni, while prompting masterclasses reached 120 Fast Stream participants. These interventions helped build both strategic leadership and practical AI skills, ensuring that senior leaders are better equipped to lead in a data- and AI-enabled environment, while future project professionals understand how to use AI tools effectively.

Alongside AI experimentation, the programme made substantial progress in improving the quality and consistency of project data. December 2025 marked a major milestone with the publication of the first-ever Project Data Standard. The standard provides a common foundation for project information across government, enabling greater consistency, transparency and interoperability. Progress has continued since publication, with implementation lead networks established across departments and industry beginning to align solutions with the standard, including the first project management and monitoring tool becoming fully aligned.

Taken together, these achievements represent far more than a collection of individual initiatives. They demonstrate the creation of a genuine ecosystem of capability: standards that improve data quality, tools that support better decision-making, learning offers that build confidence, and communities that encourage innovation and collaboration.

Yet we are clear that the journey is not over. In many respects, it is only beginning. The formal Data and AI Programme is now being closed because it has delivered on the commitments set out in the Belfast Paper and because many aspects of AI literacy and productivity are increasingly being driven by the wider system. Importantly, closure does not signal an end to the work itself. Instead, it marks a transition from programme delivery to business-as-usual adoption.

The challenge ahead is operationalisation. The foundations are now firmly in place, but the anticipated benefits will only be realised if organisations continue to embed these capabilities into everyday project delivery practice. Better outcomes depend on widespread adoption, continued learning, high-quality data, and the effective use of AI-enabled tools across the profession.

As Dame Antonia Romeo observed earlier this year: “This is happening now. We need to scale it up, talk about it, and share it.” The task for the next phase is therefore clear. We have proven what is possible. Now we must ensure these innovations become the norm rather than the exception, unlocking lasting improvements in productivity, oversight and project outcomes across government.