One in six workers in Britain is paying out of their own pocket for an artificial intelligence tool to help with their job, collectively spending nearly £1 billion ($1.4 billion) a year, according to research published by Deloitte on Wednesday.
The findings point to a widening gap between how quickly employees are adopting generative AI and how quickly employers are making the technology available in the workplace. Rather than waiting for companies to approve or provide AI tools, a significant share of workers are turning to services such as ChatGPT, Claude, Google Gemini and Microsoft Copilot themselves.
Deloitte’s inaugural GenAI Workforce Survey found that about two-thirds of workers had used generative AI tools, while nearly a quarter were using them every day as part of their jobs. More than three in 10 respondents, or 31%, said they were using AI without their employers’ knowledge, a practice commonly referred to as “shadow AI.”
Register for the next Tekedia Mini-MBA.
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
At the same time, 17% said they were personally paying for at least one generative AI tool for work.
“UK workers are showing they don’t want to wait for permission to use GenAI,” said Hayley McKelvey, Deloitte UK’s chief AI officer.
“Many are already using free tools or pay for premium versions themselves to help them get work done.”
The scale of employee-funded adoption suggests that generative AI is no longer confined to formal corporate technology programmes. Workers are increasingly treating AI assistants as ordinary productivity tools, even where their employers have not established an approved platform, policy or budget for their use.
The development has yielded a new set of questions for companies around data protection, intellectual property, security and accountability. When employees independently select AI services, organizations may have less visibility over what information is being entered into those systems and how the tools are being used.
The survey found that the most common workplace applications remain relatively straightforward. Workers primarily use AI to search for information, draft emails, and produce summaries. Those activities nevertheless generated measurable time savings. Respondents estimated that generative AI saved them an average of 70 minutes a week.
Deloitte partner Paul Lee said generative AI had rapidly become part of everyday working life, but its use remained concentrated on relatively basic tasks.
“Yet for many workers its role remains relatively basic, focused on searching for information and drafting content rather than supporting more complex work,” Lee said.
Companies have been assessing whether their growing AI investments are producing meaningful productivity gains. Widespread access to a chatbot can make individual tasks faster, but the larger economic opportunity may depend on whether businesses redesign workflows around AI rather than simply give employees another software application.
“The organizations that gain the greatest advantage from AI will be those that move beyond simply providing access to tools and focus on helping people use them with the appropriate guardrails and purpose,” Lee said.
The survey therefore captures two parallel developments. Employees are already integrating AI into their daily routines, while many organizations are still working out how the technology should be deployed, governed and connected to more complex business processes.
The £1 billion annual spending estimate also reveals an unusual feature of the current AI adoption cycle: part of the cost is being absorbed directly by workers.
Paying for a premium AI subscription, for employees, can be a relatively simple way to gain access to more capable models or additional features. However, widespread unofficial use can make it harder to determine which AI tools are being used, what value they generate and what risks they introduce.
The survey was conducted by Ipsos for Deloitte and covered 25,000 workers, giving the findings a broad view of how generative AI is entering Britain’s workplaces.
The data suggest that the next stage of workplace AI adoption may be less about persuading employees to use the technology and more about bringing existing usage into formal company systems. The challenge for employers is to turn what is currently a largely employee-driven experiment into controlled and purposeful use without removing the productivity gains that encouraged workers to adopt the tools in the first place.