Public sector AI adoption at pivotal point, new TLT report finds

Public sector leaders reveal shift in focus from technological to organisational investment as key to scaling AI across public services.

New research from UK national law firm TLT has found that public sector organisations are increasingly focused on turning AI ambition into operational reality, with leaders identifying governance, procurement, workforce and resilience capabilities as critical to deploying AI safely and at scale.

The report, Public Sector AI Perspectives: The realities of AI adoption and implementation draws on research with 100 public sector leaders across local government, central government, regulators, and blue light services.

It finds strong appetite for AI across the public sector, with leaders seeing significant potential for the technology to improve efficiency and productivity (67%), decision-making (62%), cost savings and efficiencies (60%), service delivery (56%) and citizen or service-user experience (53%) by 2030.

Confidence in implementation, however, remains low: fewer than one in ten leaders questioned feel confident their organisation can meet the legal and regulatory expectations required for successful adoption, while 45% are either not confident or unsure.

The findings come as departments are under increasing pressure to move from AI pilots to broader implementation, with the government’s AI Opportunities Action Plan calling for a “scan, pilot, scale” approach to adoption. Against that backdrop, TLT’s research highlights a “pilot plateau”, where activity and intent around AI are high but adoption has not yet translated into widespread operational deployment. The firm says the challenge is not a lack of ambition, but the need to strengthen the legal, commercial, operational and workforce structures that allow AI to move from pilots into live public service environments.

The report identifies several organisational priorities that leaders believe will be critical to scaling AI across public services:

• Addressing fragmented systems: Just over half (52%) of public sector leaders questioned say current systems do not integrate effectively across teams or departments, while 18% report no integration at all. 53% cite siloed platforms as a barrier to collaboration and efficiency with 51% pointing to a lack of shared data and 58% reporting work is duplicated often or very often.

• Governance fit for AI delivery: More than half (56%) say governance or approvals are currently a major cause of delays when implementing new policies or service improvements, with 46% identifying complex governance as a barrier to efficiency and 34% citing a lack of governance and assurance frameworks as a key obstacle to adoption.

• Strengthened AI procurement capability: 57% of public sector organisations expect to rely heavily on third-party suppliers as they build AI capability, but 53% cite limited internal expertise as a procurement challenge, 48% point to slow procurement cycles, 47% say they lack AI-specific procurement frameworks and 40% report difficulty evaluating suppliers.

• Improved workforce readiness: Two-thirds (66%) of public sector leaders questioned expect AI to bring significant or transformational change to workforce roles and skills, yet only 22% are planning leadership training for an AI-enabled workplace, focusing instead on AI literacy (62%), as 45% cite lack of skills or capacity as a main barrier to adoption.

• Increased organisational resilience: Only seven per cent of public sector leaders questioned say their organisation is well prepared to respond to AI-related failures, while 72% are not yet mostly or well prepared. The report recommends that failure readiness should be a precondition for scale, with clear thresholds, escalation routes, rollback steps and review processes defined before deployment.

• Readiness for next-generation AI: Looking ahead, only 22% of public sector organisations say they are prepared or very prepared for the shift from traditional automation to agentic AI, and just 4% describe themselves as very prepared.

Gareth Oldale, Partner and Head of Digital, Data and Commercial at TLT, said: “AI is not simply another technology wave for the public sector to adopt incrementally. It is a structural shift that will redefine accountability, decision-making and service delivery. What this report makes clear is that ambition is not the problem — readiness is. Organisations are dealing with fragmented systems, siloed data, unclear ownership and governance models that were not designed for AI-enabled delivery.

“It is clear that we are at an important moment for the public sector. The era of AI experimentation is giving way to implementation, and the next phase will be defined by execution. Ultimately, trust will be the foundation of successful AI adoption, earned through robust governance, transparency, accountability and responsible use of technology.”

Bill Hull, Partner and Head of Government and Public Services at TLT, commented: “The public sector now faces a strategic choice: treat AI as a series of pilots at the edge of the organisation or use it to fundamentally modernise how government and public services function. Those that create lasting value will be the ones that align AI to mission-critical priorities, strengthen governance early and build the legal and operational foundations for scaled adoption.

"There is clear demand for better-designed, not increased, governance, and those organisations that treat governance as an enabler of adoption rather than a compliance hurdle will be well positioned to convert ambition into scalable deployment."