Insights AI News UK GenAI workforce survey 2026: How to curb shadow AI
post

AI News

16 Sep 2026

Read 9 min

UK GenAI workforce survey 2026: How to curb shadow AI

UK GenAI workforce survey 2026 shows employers how to reduce shadow AI and protect productivity now.

The UK GenAI workforce survey 2026 shows rapid AI adoption, big personal spending, and rising shadow AI. 63% use GenAI for work, 31% do so without employer oversight, and workers spend an estimated £958m of their own money. Here is why it happens and how to curb it fast. UK workers are moving faster than their companies on AI. Many try free tools or pay for their own upgrades to save time and get work done. But this creates risk. Data can leak. Workflows can break rules. Managers feel unsure. Employees feel stigma. The UK GenAI workforce survey 2026 gives leaders a clear map: make AI safe, simple, and supported, or shadow AI will grow.

What the UK GenAI workforce survey 2026 tells us

Adoption at speed

– 63% of working adults say they knowingly use GenAI at work. – 24% use it every day. – Many rely on free tools or bring their own.

Shadow AI is real

– 31% of GenAI users use tools without employer knowledge. – 65% say leadership is not clear about how to use AI. – 23% feel stigma when they use AI at work.

Money and time

– 17% of GenAI users pay for at least one tool themselves. – Estimated personal spend reaches £958 million per year. – Average time saved is about 70 minutes per week, mostly reinvested into the same job.

Why shadow AI grows

  • Access is slow: workers wait for approvals and licenses.
  • Policies are unclear: people do not know what is allowed.
  • Stigma is real: users fear managers will think AI can replace them.
  • Training is thin: half of users report no formal guidance.
  • Personal tools feel better: premium features at home beat basic tools at work.
  • Pressure is high: deadlines push people to “just use something.”

10 practical steps to curb shadow AI without killing momentum

  • Publish a one-page AI use policy. State what’s in-bounds (use cases), what’s out (sensitive data), and who to ask. Use plain English and examples.
  • Centralize access. Buy company licenses for approved tools, enable single sign-on, and turn on enterprise controls like audit logs and retention limits.
  • Protect data by default. Apply data loss prevention, redact personal or secret info, and classify documents so models only see what they should.
  • Pick safe defaults. Offer enterprise chatbots and writing aids with content filters, watermarking, and opt-out of training on company data.
  • Train for the job. Give short, role-based lessons on prompts, fact-checking, citations, and when to keep a human in the loop.
  • Equip managers. Provide a toolkit to set team norms, review outputs, and talk openly about AI use without blame.
  • Set up governance that scales. Keep a simple AI risk register, review vendors, and tie approvals to data sensitivity and impact.
  • Measure value, not hype. Track time saved, quality gains, error rates, and rework. Share wins and lessons in weekly updates.
  • Create an AI champions network. Nominate power users in each function, run office hours, and publish vetted prompts and templates.
  • Offer a safe sandbox. Let people explore in a protected environment with sample data and clear “do/don’t” guardrails, not their personal accounts.

Use cases to move beyond basics

From quick drafts to real workflows

– Search and summarize on approved documents, with sources shown. – Meeting notes and action lists, stored in the right system. – Code suggestions tied to company repos and style rules. – Customer email drafts with tone controls and compliance checks. – First-pass analysis of spreadsheets with clear steps and logs. These moves keep the benefits of GenAI while reducing errors, leaks, and rework. They also replace ad hoc personal tools with shared, reliable workflows.

Guardrails for UK compliance

  • Follow UK GDPR and ICO guidance. Limit personal data, log processing, and run DPIAs for higher-risk use cases.
  • Use NCSC security patterns. Protect secrets, apply least privilege, and monitor for misuse.
  • Mind sector rules. FCA, NHS, and other regulators expect clear records, controls, and human oversight.
  • Review cross-border data flows and model providers’ sub-processors.
  • Keep humans accountable. AI assists; people remain responsible for decisions.

According to the UK GenAI workforce survey 2026, change must start now

The survey will run every six months. That means your progress will be measurable over time. Act before habits harden. Make approved tools easy to use. Make policies simple to follow. Make training short and useful. Reward safe wins. When people trust the path, they stop hiding their tools. Leaders can turn shadow AI into secure value if they move with clarity and speed. Provide access, set guardrails, teach skills, and model the behavior you want. The UK GenAI workforce survey 2026 is a wake-up call: workers are ready. It is time for organisations to catch up. (p)(Source: https://www.deloitte.com/uk/en/about/press-room/british-workers-spend-one-billion-pounds-of-their-own-money-on-gen-ai-for-work.html)(/p) (p)For more news: Click Here(/p)

FAQ

Q: How much are UK workers spending on GenAI for work according to the UK GenAI workforce survey 2026? A: The UK GenAI workforce survey 2026 found that UK workers are spending an estimated £958 million annually of their own money on GenAI tools used for work, with 17% of GenAI users reporting they pay for at least one tool themselves. The figure is based on responses from a representative sample of 25,000 workers surveyed by Ipsos UK between 7 May and 10 June 2026. Q: How common is GenAI use at work across the UK workforce? A: The survey found that 63% of UK working adults aged 18–70 say they knowingly use GenAI for work and that 24% use it every day. When asked about the types of tools they use, 46% report using free tools, 34% use external tools paid for by their employer and 17% use in‑house tools. Q: What does the survey mean by “shadow AI” and how prevalent is it? A: Shadow AI refers to workers using GenAI tools without their employer’s knowledge, and the survey found that nearly a third (31%) of GenAI users do this. The report also highlights that 65% of GenAI users see a lack of convincing leadership on how GenAI should be used and that 23% feel a stigma around using the tools, which can drive concealment. Q: What risks to organisations does ungoverned GenAI use create? A: The survey highlights risks including potential data leakage, workflows breaking rules, and managers feeling unsure when employees bring personal GenAI tools into work. It also found around half of GenAI users have received no formal training and that 64% of weekly users worry managers will think GenAI can do their jobs. Q: What practical steps does the article recommend to curb shadow AI without killing momentum? A: The article sets out ten practical steps such as publishing a one‑page AI use policy, centralising access and buying company licences, protecting data by default with data loss prevention and redaction, offering safe defaults and sandboxes, and providing short role‑based training and manager toolkits. These measures are intended to preserve productivity gains while reducing leaks, errors and reliance on ad hoc personal tools. Q: Which everyday tasks do employees most commonly use GenAI for and how much time does it save? A: Top reported use cases are searching for information and drafting emails (each 43%), followed by creating summaries (31%), and users say GenAI saves them about 70 minutes per week on average. The survey notes most of that saved time is typically reinvested into doing more work for the same employer. Q: What governance and compliance points should UK organisations consider when deploying GenAI? A: The survey recommends following UK GDPR and ICO guidance, applying NCSC security patterns, running DPIAs for higher‑risk use cases, and paying attention to sector rules such as those from the FCA or NHS that expect clear records and human oversight. It also advises reviewing cross‑border data flows, vetting model providers’ sub‑processors and keeping humans accountable for decisions assisted by AI. Q: How will repeating the UK GenAI workforce survey 2026 help organisations manage GenAI adoption over time? A: The UK GenAI workforce survey 2026 will be repeated approximately every six months, enabling organisations to measure how GenAI usage, shadow AI and governance evolve over time. Organisations can use the ongoing results as a benchmark to track whether interventions like approved tools, policies, training and sandboxes reduce shadow AI and increase safe adoption.

Contents