What UK leaders are really seeing in the AI transition
Over the past month, Twisted Loop has sat with leaders across eleven sectors - creative agencies, finance, construction, telecoms, not-for-profits, law and technology - across four roundtables in four weeks. What surfaced was a picture of an economy moving into its next phase of AI transition.
AI is shifting from optional tooling to the substrate of organisational design, and leaders across industries are beginning to describe the same fault lines, the same pressures and the same crossroads ahead. The story is one of cautious optimism, ambition constrained by skill and knowledge, but a growing belief that 2026 and 2027 will be decisive.
This is what leaders told us and what it means for organisations preparing to move deliberately into their next phase of AI adoption.
Data readiness is the real bottleneck
Across every roundtable, leaders unearthed the same constraint: data. Not models, not imagination, not talent - data.
Many organisations want intelligent assistants, automated workflows and autonomous systems, yet few can answer foundational questions about the data they collect, its consistency and whether it is reliable enough to power modern AI pipelines.
Take the marketing agency world. At Prolific North’s Growth to Gold event, leaders were candid about capability gaps and internal uncertainty. Most are experimenting with AI tools, yet 83% haven’t built any bespoke solutions in-house.
Aside from a small group forging ahead - productising their data and creating new revenue streams - the majority aren’t viewing data as a growth lever. They’re overlooking the opportunity to turn years of lived business knowledge buried in documents, CRMs, and siloed SaaS tools into powerful internal and client-facing products.
As the head of business intelligence at The Growth Company put it, AI remains constrained by “the infrastructure, the data, the analytics”, the foundations that still need to be in place before machine learning can scale responsibly.
The pattern is clear: data maturity is the limiting factor for every future capability - agent deployment, generative tooling, workflow automation, and the enterprise value uplift that propels businesses. Organisations that build solid foundations now will widen their strategic options dramatically.
A surge in optimism… paired with deep uncertainty
Leaders across all four events spoke with genuine excitement about what AI makes possible. Yet nearly all paired that optimism with uncertainty about how to operationalise it.
The clearest illustration came from the agency sector. Our Agency AI Adoption Survey with Mustard and Prolific North revealed that nine in ten professionals are using AI, yet only two percent feel very prepared for what comes next. Leaders described themselves as structurally underprepared even as they recognise the scale of change ahead. As Twisted Loop’s Exec Chair Jonathan Summers observed, agencies “want to do it; they’re just not doing it in a structured, deliberate way”.
Legal teams expressed a similar discomfort. Off-the-shelf generative tools behave, as one participant put it, like “a really, really rubbish trainee you have to check”, highlighting the gap between enthusiasm and trust and showcasing why developing bespoke, secure AI capability tailored to your business workflows rather than buying AI SaaS tooling can often be the best way forward for scaling businesses – especially when paired with better AI fluency for staff.
Developers are similarly split. Some are sprinting ahead, others are holding back, testing how much of their craft can be safely automated without diminishing the value they bring. The emotion underpinning all of this is not fear. It is ambiguity. Leaders understand the opportunity; they are searching for the structures that turn exploration into enterprise-level impact.
This is where accelerator-led delivery becomes essential: a way of moving quickly while building the internal capability that organisations will depend on later.
AI is reshaping organisations before it reshapes industries
Another signal stood out clearly across sectors: AI is already changing organisations long before it reshapes entire industries.
In corporate finance, one advisor predicted the traditional talent pyramid flattening into “more of a rectangular shape” - junior roles contracting and learning the ropes in new ways as AI absorbs routine work, while expectations for judgement and synthesis increase.
In agencies, many stated they’ll continue hiring early-career talent, especially into creative roles, arguing that “creativity is the last value lever we can pull”. Training will evolve, but the need for human originality will persist.
Construction leaders echoed this with examples of design automation and regulatory review: they foresee AI helping to accelerate approvals and reduce administrative bottlenecks, shifting the skill mix long before robotics alters site work.
Across all sectors, the pattern is consistent. As AI takes on more cognitive load, human value shifts toward judgement, originality, curiosity, and interpretation. These become differentiators as automation increasingly commoditises execution.
The new competitive line: Experimentation vs Capability
A clear competitive divide is emerging across UK organisations: those experimenting with AI, and those deliberately building capability.
For instance, our Agency AI Adoption Survey shows a clean distribution across three groups: non-movers, early-stage tinkerers and capability leaders. Every roundtable, regardless of sector representation, echoed this pattern.
Some founders live in the “I don’t know what I don’t know” zone. Others confessed they are “200 yards into a marathon,” aware of the distance still to run. Meanwhile, the most advanced teams are already restructuring workflows, deploying lightweight internal tools and embedding reinforcement loops that build proficiency quickly.
The difference is not enthusiasm, it’s structure. Organisations progress faster when they combine insight with accelerators - programmes that blend expert guidance with assets, templates and ready-to-go approaches that teams can start using immediately.
In a moment defined by uncertainty, leaders want progress that is both fast and safe. Twisted Loop’s approach offers exactly that: a way of embedding capability while maintaining momentum.
Why 2026–2027 is the inflection point
Over the course of the last month, Jonathan has consistently set out the macro picture with unmistakable clarity. Big tech will invest over $400bn in AI by the end of this year, rising above $570bn in 2026. Compute will reach up to 1,000× that which trained GPT-3. R&D agents will run autonomously for 8+ hours, accelerating discovery beyond traditional cycles.
Based on research evidence coming from the likes of OpenAI, Anthropic, Google and Microsoft, we predict that by the end of 2026, at least one model will match the performance of human experts across many industries and by 2027, models will frequently outperform experts on many tasks.
Leaders across sectors independently identified this window as the moment AI becomes unavoidable organisational infrastructure. Regulation will expand. Governance will tighten. Talent structures will reshape. Sector norms will reconfigure.
The horizon is no longer hypothetical - it demands preparation and deliberate capability building to be properly harnessed.
What leaders should do now: move with accelerator-led delivery
The organisations progressing fastest are those combining strategic clarity with accelerators that embed capability immediately. Leaders should focus on:
1. Building AI fluency across the workforce
The foundational skill of the next decade, understanding how to instruct, evaluate and collaborate with AI.
2. Establishing a clear AI blueprint
A shared direction that reduces uncertainty and evolves alongside the organisation.
3. Redesigning workflows for Team + Machine
Not patching legacy processes but rethinking how work flows in an environment shaped by multi-model AI, RAG systems and autonomous agents.
4. Deploying internal tools early
Even small bespoke agents build confidence, reveal opportunities and create cultural momentum.
5. Creating practical governance
Guardrails that enable responsible speed, not frameworks that slow teams down.
6. Preparing for talent shifts
Revisiting training, recruitment and role design early will prevent future pressure from becoming acute.
Accelerator-led delivery supports all of these steps. It allows leaders to move quickly while still building long-term scaffolding for capability, ensuring that progress is both rapid and sustainable.
Moving deliberately into the Team + Machine era
Across four weeks and eleven sectors, the same story resurfaced repeatedly. Leaders are optimistic, ambitious and acutely aware of the practical barriers ahead. The convergence across industries is unmistakable: AI is no longer an optional accelerator. It is becoming the infrastructure of competitive advantage.
The organisations that combine strategic intent with accelerator-led delivery will find that capability compounds quickly. Those that wait may discover that the operating assumptions of their sector have already moved on.
The next step is to begin confidently, harnessing the right accelerators to propel progress.
How twisted loop can help your business
Our team has been delivering data and AI solutions for over 30 years.
We know how to start fast, align strategy to business goals, and build momentum with confidence.
We offer a range of AI Strategy and Rapid Prototyping services, with delivery timescales as short as two weeks - so you can choose the balance of speed, rigour, and scope that best fits your business priorities.
Want to rapidly accelerate your AI strategy or test a high-impact use case?
