Will AI Replace Your Job? Which UK Careers Are Most (and Least) at Risk

The conversation about AI and jobs has moved from speculation to measurable impact. In 2026, UK employers are actively restructuring workforces around AI capabilities, government reports are quantifying displacement risk, and the labour market data is beginning to show real shifts. Here is what the evidence says about which UK careers are most and least at risk, and what the numbers mean for your own career planning.

The Scale of the Threat

The Institute for Public Policy Research (IPPR) estimates that up to 8 million UK jobs could be affected by AI if adoption focuses purely on cost-cutting automation without policy intervention. That is roughly a quarter of the UK workforce. The figure is not a prediction — it is an upper bound under a worst-case scenario — but it signals the potential magnitude.

More immediately, a 2025 survey of UK executives found that 65% planned to reduce headcount before the end of 2026, citing AI-driven efficiency gains as the primary driver. A further 8% said they would freeze recruitment entirely. These are not hypothetical projections — they are stated intentions from people making hiring decisions right now.

Which Roles Are Most Exposed?

AI exposure is not evenly distributed across the economy. Research from the UK government’s Department for Science, Innovation and Technology, combined with ONS labour market data, allows us to rank broad categories of work by their vulnerability.

High Risk: Administrative and Back-Office Roles

The roles most immediately affected are those involving routine information processing. These include:

• Data entry clerks: £18,000-£24,000. Large language models and document processing AI can now handle the majority of structured data entry tasks. Vacancy volumes have fallen approximately 30% since 2023.
• Bookkeepers and payroll administrators: £22,000-£30,000. Automated accounting platforms have steadily reduced demand. AI-powered tools now handle invoice matching, reconciliation, and basic payroll processing with minimal human oversight.
• Customer service representatives (text-based): £20,000-£26,000. AI chatbots and automated response systems handle an increasing share of tier-one support queries. Companies report 40-60% deflection rates on routine enquiries.
• Legal secretaries and paralegal assistants: £22,000-£32,000. Document review, contract summarisation, and legal research — previously time-intensive human tasks — are increasingly handled by AI tools.
• HR administrators: £23,000-£30,000. Nearly half of executives surveyed said AI agents can already perform more than 50% of a current HR professional’s administrative duties, including screening, scheduling, and compliance documentation.

The common thread is predictable: roles centred on processing, sorting, and summarising information that follows established patterns.

Medium Risk: Professional Knowledge Work

A second tier of roles faces partial disruption rather than wholesale replacement. These are jobs where AI augments human capability but cannot yet replace the judgement, client relationships, or creative elements involved.

• Junior software developers: £28,000-£38,000. This is perhaps the most discussed category. AI code generation tools have measurably increased senior developer productivity, reducing demand for junior developers who previously handled routine coding tasks. ONS data shows a striking 44% drop in the number of 16-24 year olds employed in computer programming in a single year.
• Financial analysts (routine reporting): £30,000-£45,000. AI can now generate standard financial reports, build models from templates, and process market data faster than human analysts. The strategic interpretation layer remains human, but the volume of analysts needed for data processing has declined.
• Content writers (commodity): £22,000-£35,000. AI generates serviceable first drafts for product descriptions, basic news summaries, and template-driven content. Writers who bring specialist knowledge, original research, or a distinctive voice retain their value. Those producing generic content at volume face direct competition from AI tools.
• Translators and interpreters (written): £25,000-£40,000. Machine translation quality has improved dramatically. Routine document translation is increasingly automated, though nuanced, culturally sensitive, or legally binding translation still requires human expertise.

Lower Risk: Roles Requiring Physical Presence, Empathy, or Novel Problem-Solving

Some careers remain largely insulated from AI displacement, at least on the current technology trajectory.

• Healthcare professionals (nurses, doctors, therapists): £30,000-£100,000+. Diagnosis support AI exists, but the physical, emotional, and ethical dimensions of healthcare require human presence. The NHS vacancy crisis ensures sustained demand regardless of AI adoption.
• Skilled trades (electricians, plumbers, heat pump installers): £30,000-£55,000+. AI cannot install a boiler, rewire a house, or fix a burst pipe. The net zero transition is actively increasing demand for these roles.
• Social workers and care professionals: £30,000-£48,000. The relational, safeguarding, and crisis-management aspects of social work are fundamentally human. Caseloads are rising, not falling.
• Senior engineers and architects: £60,000-£120,000+. Complex system design, novel problem-solving, and cross-disciplinary judgement remain beyond AI capabilities. These professionals increasingly use AI as a tool, making them more productive rather than redundant.
• Teachers and lecturers: £30,000-£55,000. Despite AI tutoring tools, the classroom management, pastoral care, and adaptive teaching elements of education require human presence. Teacher shortages persist across the UK.
• Construction project managers: £40,000-£90,000. Coordinating physical construction sites, managing subcontractors, and navigating real-world logistics involves too many unpredictable variables for AI automation.

The Emerging AI Job Market

The displacement story is only half the picture. AI is simultaneously creating new roles and expanding existing ones.

UK job postings mentioning AI or related tools reached 5.6% of all listings — the highest share to date and above peer economies including France, Germany, the US, Canada, and Australia. The roles driving this demand include:

• AI/ML engineers: £65,000-£120,000+. Building, training, and deploying machine learning models. Demand has grown 35-40% year-on-year since 2023.
• AI product managers: £55,000-£90,000. Translating AI capabilities into viable products. A new role category that barely existed three years ago.
• Data engineers: £45,000-£75,000. Building the data pipelines that AI systems depend on. Demand has grown steadily and shows no sign of slowing.
• AI safety and governance specialists: £50,000-£85,000. As regulation increases, companies need people who understand both the technology and the compliance landscape.
• Prompt engineers and AI integration specialists: £35,000-£65,000. Optimising AI tool usage within existing business processes. An entirely new job category with rapidly formalising skill requirements.

The Junior vs Senior Divide

One of the most consequential patterns in the data is the widening gap between junior and senior professionals. AI tools are most effective at automating entry-level tasks — the very tasks that have traditionally served as training grounds for new professionals.

This creates a structural problem. If companies hire fewer juniors because AI handles their traditional workload, the pipeline of experienced professionals shrinks. The short-term cost saving creates a long-term talent shortage. Some sectors are already seeing this dynamic play out: law firms reporting difficulty finding mid-level associates because they hired fewer trainees during the initial AI adoption wave.

For individuals entering the workforce, the implication is clear. Entry-level roles that consist primarily of processing and summarising are increasingly poor bets. Entry-level roles that involve client interaction, physical presence, creative judgement, or working alongside AI tools are more resilient.

What the Government Data Actually Shows

The UK government’s assessment of AI capabilities and labour market impact, published by the Department for Science, Innovation and Technology, provides the most rigorous domestic analysis. Key findings include:

• Approximately 10-30% of UK jobs are highly exposed to AI, meaning a significant portion of their tasks could theoretically be automated with current technology.
• Exposure does not equal displacement. Many highly exposed roles will be augmented rather than replaced, with AI handling specific tasks while humans focus on higher-value work.
• The transition will be gradual. Even optimistic adoption scenarios project displacement occurring over years, not months, giving workers time to adapt if they act proactively.
• Policy matters enormously. Countries that invest in retraining and manage the transition actively will see different outcomes from those that leave it to market forces.

What Should You Do?

The data points to several practical conclusions.

If you are in a high-risk role: Start building adjacent skills now. A bookkeeper who learns financial analysis and advisory work is far more resilient than one who only processes transactions. A customer service representative who develops expertise in complex complaint resolution and client retention has value that chatbots cannot replicate.

If you are in a medium-risk role: Learn to use AI tools effectively within your profession. A junior developer who can leverage AI to work at a senior level of productivity is more valuable than one who competes with AI on basic coding. The professionals who thrive will be those who treat AI as a force multiplier, not a threat.

If you are entering the workforce: Prioritise roles that combine technical literacy with human skills — empathy, complex communication, physical dexterity, or novel problem-solving. The safest careers are not necessarily the most technical, but the ones where human presence and judgement remain essential.

If you are in a low-risk role: Do not be complacent, but recognise your relative advantage. A nurse, electrician, or social worker is not immune to AI-driven changes in their work, but the core of what they do remains fundamentally human. Focus on staying current within your profession rather than pivoting unnecessarily.

The evidence suggests that AI will transform the UK labour market significantly but not catastrophically. The workers most at risk are those in routine information-processing roles who do not adapt. The workers best positioned are those who combine domain expertise with the ability to work alongside AI tools — regardless of which sector they are in.