Skip to main content
OpenRole
All posts
Research17 March 2026·8 min read

Which Industries Does AI Recommend Working In? The UK Data

Which Industries Does AI Recommend Working In? The UK Data

When candidates ask AI for career advice, it doesn't just answer — it recommends.

"Which industries are good to work in?" is one of the fastest-growing query types in AI career research. We tested what ChatGPT, Perplexity, and Google AI recommend to UK job seekers — and compared their suggestions against actual employment data, salary growth, and job satisfaction surveys.

The findings reveal significant AI biases that favour some industries and unfairly penalise others.

What we tested

We asked three AI models a series of career direction questions:

  1. "Which industries pay the best in the UK?"
  2. "What are the best industries to work in the UK in 2026?"
  3. "Which UK industries have the best work-life balance?"
  4. "What industries are growing fastest in the UK?"

Each question was asked five times with slight variations to check consistency.

AI's recommendations vs reality

AI's "best industries" (aggregated across all models)

  1. Technology — Recommended by all three models in all variations
  2. Financial services — Consistently recommended for compensation
  3. Healthcare — Recommended for job security and purpose
  4. Renewable energy — Recommended for growth and purpose
  5. Professional services — Occasionally recommended

What AI consistently undervalues

  1. Manufacturing — Rarely mentioned despite £35K+ average salaries and growing automation roles
  2. Construction — Almost never recommended despite severe labour shortages driving up pay
  3. Logistics and supply chain — Absent from recommendations despite being one of the UK's largest employers
  4. Public sector (non-NHS) — Dismissed as low-paying despite pension value equivalent to £8-12K additional salary
  5. Hospitality — Actively discouraged in most AI responses

The tech bias

Technology is AI's overwhelming recommendation across all models and all question variations. The reasoning is typically:

  • "High salaries and strong salary growth"
  • "Remote working opportunities"
  • "Innovation and exciting work"
  • "Strong demand for talent"

This isn't wrong, but it's incomplete. AI significantly overstates tech sector stability (ignoring the 2023-2024 layoff cycles), understates the competitive entry requirements, and presents tech salaries as representative when they're actually skewed by a small number of very high-paying companies.

The data AI misses:

  • UK tech salaries outside London average £45K-£55K — competitive but not the outlier AI suggests
  • Tech sector job satisfaction ranks 4th in the UK, behind education, charity/non-profit, and healthcare (CIPD data)
  • The UK tech sector added 80,000 jobs in 2025, but manufacturing added 45,000 — both are growing, but AI only mentions one

The public sector penalty

The most striking bias was against the public sector.

When asked about public sector careers, AI responses included:

  • "Salaries are significantly below private sector equivalents"
  • "Career progression is slow and bureaucratic"
  • "Budget constraints limit resources and innovation"

What AI ignores:

  • The NHS pension scheme alone is worth £8,000-£12,000 in equivalent salary annually
  • Local government defined benefit pensions add similar value
  • Public sector annual leave is typically 27-33 days vs 25 in private sector
  • Flexi-time and compressed hours are widely available
  • Job security is measurably higher

When you include pension value and additional leave, a public sector role at £35K often has total compensation equivalent to a £45K-£50K private sector role. AI never makes this calculation.

Why AI is biased

The bias isn't intentional — it's structural:

1. Training data skew

AI models are trained disproportionately on English-language internet content, which over-represents tech, finance, and professional services. These industries produce more blogs, more Reddit discussions, more career advice content. Manufacturing workers don't typically write Medium posts about their career paths.

2. The Reddit effect

Reddit is the #1 or #2 cited source for all major AI models on employer topics. Reddit's demographics skew young, urban, and tech-adjacent. The industries Reddit users discuss favourably are the industries AI recommends.

3. Recency bias

AI models tend to cite recent, prominent content. Tech companies dominate recent career content through engineering blogs, open-source communities, and social media presence. Traditional industries have less digital footprint.

4. Salary data availability

Tech and finance companies more frequently publish salary data (through job listings, transparency reports, and platforms like Levels.fyi). AI has better data for these sectors, which means it can make more specific recommendations — creating a confidence that reads as endorsement.

What this means for employers in "unfavoured" industries

If you're a manufacturing company, a construction firm, a local council, or a hospitality business — AI is actively steering candidates away from your industry. Not with malice, but with ignorance.

The practical impact:

  • Candidates who ask "should I work in manufacturing?" get lukewarm or negative responses
  • Industry-wide, this compounds into a talent pipeline problem
  • The candidates you most want — talented people with options — are being told to go elsewhere

How to fight industry bias

For individual employers:

  1. Publish your actual compensation data (including pension value and total package) in machine-readable formats. Force AI to recalculate.
  2. Create content that challenges the narrative. "Why I chose manufacturing over tech" or "What my pension is actually worth" — publish it on your domain where AI can cite it.
  3. Deploy llms.txt with the full picture. Include total compensation, not just base salary. Include career progression data. Include job satisfaction metrics.

For industry bodies:

  1. Commission and publish AI-specific research. "The Real Value of a Public Sector Career" backed by compensation analysis — this becomes a citation source.
  2. Create industry-wide salary transparency initiatives. The sectors with the most published data get the best AI recommendations.
  3. Engage with AI companies directly. AI models have feedback mechanisms. Report inaccurate industry characterisations.

The candidate perspective

If you're a candidate reading this: AI career advice is biased. Use it as one input, not the input.

Before writing off an industry based on what ChatGPT said:

  • Calculate total compensation (salary + pension value + benefits + leave)
  • Look at actual employee satisfaction data (CIPD, Best Companies)
  • Talk to people who actually work in the sector
  • Consider your personal priorities — stability, purpose, flexibility, progression

AI optimises for the average of its training data. Your career isn't an average.

See what AI says about your company

Whether you're in a "favoured" or "unfavoured" industry, AI has a story about your company. Run a free audit to see what candidates actually hear.

30 seconds. No signup.

→ Run your free AI employer brand audit

Sources: OpenRole research (March 2026), Profound — 680M LLM citation analysis (2025), CIPD UK Working Lives Survey (2025), ONS Annual Survey of Hours and Earnings (2025), UK Government sector employment data (2025).