HR teams are running more programmes, more communications, and more individual cases than headcount has grown to support. AI is the realistic answer to volume work (drafting, analysis, benchmarking) when it is applied with the right governance for sensitive data. Five practical use cases below, with the prompts to copy.
1. Job description builder with DEI audit built in
Tool: Claude
Two prompts, one strong inclusive JD:
Write a job description for a Senior Workplace Experience Manager. Include: role summary, five key responsibilities, required qualifications, and preferred experience.
Then:
Audit the description above for language that may unintentionally discourage applications from women, neurodivergent candidates, or non-traditional career paths. Rewrite any flagged phrases.
2. Policy update communications
Tool: Claude Projects
Upload your existing hybrid work policy:
This policy has been updated to change the minimum in-office requirement from 2 to 3 days per week. Write: a manager briefing note (200 words), an all-staff email (150 words), and a FAQ of five questions employees are likely to ask, with answers.
Three documents in five minutes.
3. Employee survey analysis
Tool: Claude
Paste anonymised survey free-text responses:
Analyse these employee survey comments. Identify: the top three positive themes, the top three concern themes, any specific operational issues mentioned more than twice, and a recommended action for each concern theme. Present as a summary table.
Qualitative insight from 200 responses in under two minutes.
4. Hybrid attendance policy benchmarking
Tool: Claude with web search or Perplexity
Research current hybrid work attendance policies across financial services firms in the UK with 500-5,000 employees. Summarise: the most common minimum in-office day requirements, how policies are enforced, and how firms are handling exceptions. Cite your sources.
Policy benchmarking that used to need an HR consultant.
5. Onboarding flow builder
Tool: Claude Projects
Upload your current onboarding checklist, IT setup process, and office access procedure:
Design a 30-day onboarding journey for a new hybrid employee joining our [city] office. Week 1 covers logistics and setup. Week 2 covers team integration. Weeks 3-4 cover role ramp-up. Include: daily touchpoints, who owns each step, and the five most common friction points new starters report.
What is safe in public AI tools, and what is not
HR data safety: what to anonymise and what to leave out entirely
| Data type | Public AI (ChatGPT free) | Enterprise AI (zero retention) | Verdict |
|---|
| Anonymised survey free-text | OK with care | Safe | Use freely |
| Job descriptions (no names) | Safe | Safe | Use freely |
| Policy text (no employee data) | Safe | Safe | Use freely |
| Named performance notes | Never | With caution | Avoid |
| Salary or comp data | Never | Never | Avoid |
| Medical or sickness records | Never | Never | Avoid |
| ER case notes with names | Never | Never | Avoid |
Risks every HR team must address
- AI in hiring decisions. Using AI to screen CVs or score candidates introduces bias risk and potential discrimination liability. AI should assist the human reviewer, not replace the decision. The EHRC has guidance on this.
- Employee data in AI tools. HR data is among the most sensitive in any organisation. Name, role, salary, performance notes, and medical information should never enter a public AI tool. Use anonymised summaries only.
- AI-drafted policy is not legal advice. AI produces plausible employment policy language. It does not know your jurisdiction, your collective agreements, or your specific legal exposure. Always have employment counsel review.
HybridHero for HR teamsThe attendance and workplace data your HR team needs.
HybridHero gives HR teams visibility on attendance, policy compliance, and workforce patterns. That is the data your hybrid policy needs to actually mean something. Already on another platform? The Switch Programme migrates you in 30 days.
Book a demo