New findings from Asana’s 2025 Global State of AI at Work Report reveal that UK workers are using AI to automate outdated systems, leading to work that lacks depth, accuracy, and critical human insight. The result is a growing “workslop economy”, where automation is creating inefficiency instead of reducing it.

The efficiency illusion: UK workers aren’t getting smarter with AI

With 59% of UK desk workers using AI weekly, the country’s workforce is among the most active adopters. Yet, only 19% understand how to use AI agents effectively, placing the UK among the least prepared nations globally.

The report shows that the UK’s employees are working harder, not smarter. They save just five hours per week using AI (the lowest figure among surveyed countries), while digital exhaustion has surged by 11% in the past year. This suggests organisations are automating broken workflows rather than redesigning how work is done, one of the key reasons workslop continues to spread.

How workslop takes hold: automation without transformation

Only 10% of UK organisations have redesigned workflows for AI, and just 21% have reached scaling maturity, leaving many stuck in pilot phases. Lack of structure is creating confusion over accountability. When AI agents make mistakes, 39% of UK workers say they don’t know who’s responsible. Lack of AI accountability allows workslop to thrive where work looks complete but is riddled with errors or missing context.

Workers describe AI Agents as inconsistent, overly confident, and prone to error, with over half reporting that AI agents ignore feedback or share incorrect information. Instead of saving time, over half (58%) of workers believe poorly performing agents actually create extra work.

Training and trust: the antidote to workslop

The report also exposes a growing training gap. While 82% of employees say proper training is essential to using AI agents effectively, only 38% of organisations provide it. Without education or boundaries, workslop breeds as teams struggle to manage errors, provide oversight, or collaborate confidently.

Despite these concerns, 53% of UK workers are enthusiastic about using AI at work and 76% of workers globally remain optimistic about AI’s potential to reshape how work gets done, but only if organisations provide the governance, training, and accountability needed to make agents trustworthy teammates rather than a source of workslop.

“Without clear ownership, training, and quality standards, AI becomes an undependable co-worker that adds noise instead of value,” said Dr Mark Hoffman, Head of AI Insights, Asana Work Innovation Lab. “To move beyond ‘workslop,’ organisations need to focus on structure: clear usage guidelines that empower safe experimentation, regular training, defined accountability when agents go wrong, and a culture that values accuracy over speed. Without those foundations, organisations risk building up ‘AI debt’, an inefficiency that compounds over time.”

For more startup news, check out the other articles on the website, and subscribe to the magazine for free. Listen to The Cereal Entrepreneur podcast for more interviews with entrepreneurs and big-hitters in the startup ecosystem.