UK Workers Spend About $1.3 Billion a Year on Generative AI for Work — Often Out of Pocket

# What the survey measured Deloitte polled 25,000 workers in the UK in what the firm calls the largest single-country study of workplace generative AI use. The survey reports fast adoption: 63% of working adults said they actively use GenAI for work. Nearly half of those users rely on free tools, while 34% use generative AI provided by their employer.

# How much employees are paying One in six workplace AI users said they paid for at least one tool themselves. Deloitte's estimate for the UK is roughly $1.3 billion per year that workers collectively spend on generative AI for their jobs, without expensing it. That figure captures subscriptions or other direct spending by staff who want AI tools but don't have access through employer channels.

# What people actually use GenAI for According to the survey, usage is concentrated in straightforward tasks. Common activities include drafting emails and using GenAI as a search replacement. Deloitte characterizes this as "relatively basic" use — workers are turning to AI for drafting and information retrieval rather than more complex, integrated workflows or decision support.

Paul Lee, Deloitte UK partner and head of industry insight, said the striking part is how quickly GenAI has become part of everyday working life. He also noted that, for many workers, GenAI's role remains basic and that workers are using it despite limited training, patchy guidance, and in some cases without their employer's knowledge.

# How the UK compares to the US The article references a roughly contemporaneous Gallup survey of US workers that shows 52% of American workers use AI in some capacity, though only 15% use it daily. While polling methods differ and aren't directly comparable, the data suggest higher reported GenAI adoption in the UK sample.

# Risks and organizational gaps The Deloitte findings highlight four practical gaps employers face:

• Cost shifting: Employees are buying tools themselves when employer-provided options are unavailable or inadequate.
• Limited capability: Current use cases are shallow, which reduces potential productivity gains and can increase error risk if staff rely on GenAI for tasks they haven't been trained on.
• Training shortfall: Many workers report limited or no formal training on how to use GenAI safely and effectively.
• Visibility and governance: Some employees use tools without informing their employer, which raises data security, compliance, and quality-control concerns.

# What employers should consider now Employers that want to manage cost, risk, and value should consider four concrete actions:

• Audit actual GenAI use across teams to understand which tools employees are buying or using informally.
• Provide approved tools or subscriptions to reduce out-of-pocket spending and to centralize security and compliance controls.
• Create short, task-focused training and clear guidance about acceptable use, data handling, and verification practices for AI outputs.

# Bottom line GenAI is already embedded in many UK workers' daily routines, but current usage is mostly basic and unevenly supported. The combination of employee out-of-pocket spending and minimal employer oversight points to gaps companies can address to contain costs, manage risk, and improve how AI actually helps get work done.