This piece draws on "The UK's AI Plan Shows Progress on Paper, But What About in Practice?", an article by our co-founder Tarek Nseir for the Forbes Technology Council, March 2026.
On paper, the UK's AI Opportunities Action Plan looks close to finished. In practice, adoption across UK business still feels slow, and the gap between the two is the real story. Our co-founder Tarek Nseir set this out in Forbes. The government counts around three-quarters of its pledges as delivered, yet delivering a plan is not the same as making an impact. When success is measured by boxes ticked rather than systems running in production, it is easy to mistake activity for achievement.
How complete is the UK's AI plan, really?
The headline number is 75%, the share of the plan's pledges the government counts as delivered. Look underneath it and the picture softens. Private-sector adoption targets, arguably the ones that matter most for growth, sit at around 67%. And some of the flagship commitments have little to show. The government signed a memorandum of understanding with OpenAI in July 2025 and presented it as a defining moment for public service reform. A Freedom of Information request by Valliance found that eight months on, not a single trial had taken place under it, with the Department for Science, Innovation and Technology confirming it had "not undertaken any trials under the memorandum of understanding with OpenAI". Steps have been taken elsewhere, but few of them bring enterprises closer to using AI at scale.
Why is enterprise AI adoption still stuck?
Most organisations are still running experiments rather than production systems. Too many are stuck in what Tarek calls perpetual pilot mode, where a project proves a point and then stalls on the way from proof of concept to deployment. Our own research shows only half of the largest tech spenders' AI projects meet their success metrics, so the constraint is not appetite or budget. Part of it is incentives. At the policy level there is still little that actively rewards businesses for adopting and scaling AI, and generic training schemes tend to measure skills gaps without closing them.
What would actually close the delivery gap?
On the government's side, that means targeted funding and procurement reform, with incentives that reward adoption at scale rather than plans on paper. On the business side, three things make the difference. Make upskilling a real priority and give people time to experiment. Focus on the AI initiatives that drive the most organisational impact rather than spreading effort thinly. And move decisively towards deployment instead of lingering in pilots. The foundations matter, but the longer we spend laying them, the less time is left to build on them.
"When success narratives rest on ticking boxes rather than real-world deployment, we risk mistaking motion for momentum." — Tarek Nseir, Forbes Technology Council
This piece draws on Tarek Nseir's Forbes Technology Council article, "The UK's AI Plan Shows Progress on Paper, But What About in Practice?".
