The next phase of UK AI adoption isn’t about speed

The rapid early adoption of tools like ChatGPT was exciting. Their value was immediate and personal, supporting the speed of task completion, summarisation, problem-solving, and drafting.

It democratised AI, reducing the distance between “AI as a concept” and “AI as something every individual can use.” This gave people a visceral feeling of the future arriving.

We’re now moving from the “excitement” phase to the “industrialisation” phase of the AI hype cycle—which feels different. As a technology and digital officer, one of the most common viewpoints I hear from leaders across the UK and Ireland (UK&I) is that AI deployment isn’t moving fast enough.

That’s because speed isn’t at the heart of what’s happening in this second stage of UK AI adoption. For the UK&I organisations I work with in regulated industries, it takes longer.

Why does AI adoption feel slower?

In the UK&I, there are four structural reasons for AI adoption feeling slower. They’re largely focused on ensuring scalability aligns with realistic capabilities.

1. Trust, risk, and regulation need to be factored into AI deployments. Regulatory boards are rightly focused on model risk, data leakage, security, resilience, and compliance. UK sector regulators and internal risk functions are sending a single message for organisations deploying AI: Prove to us that it’s safe, controlled, and accountable.

2. Data quality needs improving. AI outcomes are dependent on good data quality and governance, but many organisations have realised they have poor-quality data estates. Inconsistent taxonomies, manual processes, and duplicate records can’t be fixed overnight.

3. Enterprise operating models haven’t caught up. AI is changing how work gets done, necessitating redefining roles, updating policies, retraining teams, and redesigning processes. For many enterprises, the barrier is their capacity for change.

4. The chief financial officer is asking more questions about return on investment. Where is AI investment saving cost, reducing risk, improving service, or generating revenue? This is pushing organisations towards fewer experiments and more scalable programmes.

Getting past the bottleneck requires focus

Another reason the UK’s AI adoption curve looks flatter than expected is that organisations are now doing the hard part. As they enter the build-out phase, AI is moving from being a destination to a critical layer across the business: a layer in workflows, decisions, service delivery, and operations.

The organisations making the most progress in AI aren’t the ones running the most pilots. Instead, they’re:

• Mapping where AI can remove cost, accelerate service, and reduce risk
• Redesigning end-to-end processes, not just bolting on AI
• Building governance models that scale across policy, risk, audit, and model lifecycles
• Aligning data, security, and architecture
• Creating a workforce adoption strategy that’s credible and humane

The pattern of AI adoption is different

There’s been no shortage of tech cycles over the years. Many followed the same pattern: excitement, inflated expectations, disappointment, and consolidation. AI is different for several reasons.

First, its benefit can be immediate and universal. AI can help improve knowledge and work across every function. In customer service, for example, service provider Kainos uses AI to generate knowledge articles so that customers can find answers faster, reducing their average time to resolve by 71%.

In HR, pharmaceutical firm AstraZeneca is using the ServiceNow AI Platform to help recruit more than 20,000 new colleagues by 2030. The organisation is freeing thousands of HR hours by:

• Automating repetitive, manual tasks
• Enabling productivity from day 1
• Creating a personalised portal for each new hire

Models are becoming more accessible, commoditised, and embedded in platforms. The question leaders are asking has shifted from “Can we access AI?” to “Can we operationalise AI safely at scale?”

AI is now being implemented to take tangible action across the workflows that run the business. Genuine build-out is happening in agentic experiences, automation, orchestration, governance, and integration into enterprise platforms.

What should UK&I leaders prioritise?

For leaders working in technology, operations, risk, or transformation and looking to drive growth with AI, here are three pieces of advice:

1. Embed AI into every workflow. Move your business to a company-wide mindset where AI is viewed as an essential element of every workflow, not an add-on or potential consideration.
2. Build an internal AI governance model designed for speed. Enable more effective adoption and risk management by defining clear patterns, approvals, controls, and ownership for your organisation
3. Treat data and process as a top order of business. Fix what’s structurally blocking value, such as inconsistent taxonomies, fragmented records, manual hand-overs, and unclear accountability.

The tech industry may have expected AI to move at the speed of consumer adoption, but enterprise AI adoption—especially in the UK&I—moves at the speed of trust, governance, data readiness, and operating model change.

Find out how ServiceNow can help you put AI to work for people.