UK firms shift AI focus from pilots to trust and results

TL;DR:

• Senior technology executives say enterprise AI is entering a phase defined by trust, integration and measurable results rather than experimentation.
• One survey found 89% of executives believe AI improves productivity, yet the 4.6 hours a week it saves is nearly cancelled out by time spent checking its output.
• The harder question, they argue, is no longer where AI can be deployed but where it can actually be relied upon.

The corporate conversation about artificial intelligence is shifting from whether to adopt it to whether it can be trusted in core operations, according to executives from five technology firms surveyed by IT Brief UK. The common thread: early enthusiasm for generative tools is giving way to a demand for dependable, production-grade systems that reduce workloads without adding new risk.

The trust gap

Dharam Gurbani, chief growth officer at Ascendion, said UK enterprises have “stopped asking whether AI can be adopted” and are instead asking where it improves productivity, eases operational drag and strengthens compliance “without adding more pressure on already stretched systems”. The most credible programmes, he argued, keep accountability with people while letting teams move faster. The numbers expose the tension. Research cited by document-software firm Foxit found that while 89% of executives believe AI is improving productivity, the average 4.6 hours a week it saves is almost entirely consumed by verifying its output. “Without confidence in the output, productivity gains quickly disappear,” said Foxit’s Evan Reiss, framing trust as “the defining competitive advantage” as AI agents take on more complex work. Workato’s Derek Thompson added that many deployments remain “fringe experiments”, summarising research or rewriting emails, leaving deeper business value untapped.

Looking forward

The framing matters for UK businesses weighing where to commit AI budgets in 2026. It echoes a growing body of evidence that trust, not technology, is the barrier to scaling AI, and it points to a maturing market in which vendors will be judged on governance, verification and measurable outcomes rather than headline capability. Verification overhead is emerging as the hidden tax on AI productivity; firms that get the data quality and assurance foundations right stand to convert pilots into genuine returns, while those chasing tool counts risk staying stuck in the experimentation phase.