Free Competency Framework Template | Role-Based Examples & Proficiency Levels
Free Competency Framework Template | Role-Based Examples & Proficiency Levels
AI interview questions for HR and People roles help hiring panels assess whether a candidate uses AI safely, practically, and compliantly across recruiting, performance, and skills workflows. Good questions don't test tool knowledge — they test judgment: when does AI help? When is a human decision mandatory? What must never go into a prompt?
Why AI competency in HR roles matters in 2026
HR teams are rolling out AI-assisted tools for CV screening, performance analytics, skills frameworks, and employee surveys. The people who buy, configure, and use these tools carry real responsibility — for data privacy, fairness, and regulatory compliance.
The EU AI Act entered into force in August 2024. AI systems used in HR — including candidate screening, performance evaluation, and internal mobility — are classified as high-risk under Annex III of the EU AI Act. Full compliance obligations apply from August 2026. That means HR candidates today must understand what that classification means for their day-to-day work.
In Germany and Austria, the BetrVG adds another layer: under § 87(1)(6) BetrVG, works councils have co-determination rights over any technical system capable of monitoring employee behavior or performance — regardless of whether that's the intended purpose. AI tools in HR regularly fall within this scope.
Organizations hiring for HR roles need candidates who don't just work around these requirements, but actively navigate them.
The 5 competency domains every HR panel should assess
AI competency in HR is not a single skill. It spans five domains that weigh differently depending on the role:
| Domain | What it covers | Most relevant for |
|---|---|---|
| 1. Practical use | Can the person embed AI into real workflows? Drafting, summarizing, templates — with clear boundaries on what AI decides. | Recruiters, HR Ops |
| 2. Data and privacy awareness | Does the person know what can go into a prompt and what can't? How do they protect employee and candidate data from AI tools? | All HR roles |
| 3. Bias and fairness awareness | Can the person spot discriminatory patterns in AI outputs and correct them? Do they understand how training data shapes results? | Recruiters, HRBPs |
| 4. Compliance and governance | Does the person know the relevant regulatory frameworks (GDPR, EU AI Act, BetrVG) and act accordingly? | HRBPs, Heads of People |
| 5. Change and enablement | Can the person manage AI rollouts, address skepticism, and bring works councils and employees along? | Heads of People, CHROs |
Questions by role: Recruiter, HRBP, Head of People
Not every AI question fits every role. A junior recruiter needs different AI skills than a Head of People with governance accountability. The questions below are organized by typical role clusters.
For Recruiters and HR Ops
These roles use AI most often for operational tasks: job ads, candidate outreach, CV pre-screening support. The key competencies to probe are judgment, data hygiene, and knowing when AI assistance crosses a line.
- Which AI tools do you use today for job ads or candidate communications — and what do you check in the output before you use it? Listen for: which tool, for which task, and what the review step looks like.
- Imagine you're using ChatGPT to help create a shortlist from 80 CVs. What would stop you from doing that — and why? A strong answer names GDPR consent, lack of bias controls, and the question of whether that system would qualify as high-risk AI under the EU AI Act.
- How do you ensure AI-assisted candidate outreach doesn't feel like spam — and doesn't expose personal data to external systems?
- A colleague sends you a finished job ad that ChatGPT wrote. What do you review before posting it? Look for: exclusionary language, inflated requirements, tone, any legal disclosures needed.
For HRBPs and People Partners
HRBPs sit between operations and business. They mediate performance processes, advise managers, and are often the first HR contact when a business unit wants to introduce a new AI tool.
- A line manager wants to use AI to evaluate team member performance on a weekly basis. How do you respond? A strong answer references § 87(1)(6) BetrVG (works council co-determination), GDPR impact assessment requirements, and the risk of algorithmic bias replacing managerial accountability.
- Where would you use AI in a performance review cycle — and where is the hard boundary for human decision-making?
- An employee asks you directly whether AI influenced their performance rating. What do you say? This tests transparency obligation awareness — do they know what must be disclosed?
- You're rolling out a skills framework with AI support. How do you ensure the competency categories aren't discriminatory?
For Heads of People and CHROs
At this level, the questions shift to governance, vendor management, and the strategic question: which AI systems do we deploy — and under what conditions?
- How would you ensure that AI-assisted HR tools in your organization meet EU AI Act requirements? Expected topics: high-risk classification audit, human oversight documentation, works council involvement, data protection officer coordination.
- A vendor tells you their ATS is "GDPR-compliant." What do you ask next? Stronger answers go further: data processing agreement, sub-processors, model training on customer data, deletion timelines, data subject request handling.
- How would you involve the works council in introducing an AI-based recruiting tool — and what would a works agreement need to cover?
- What would your first step be if you discovered that an HR tool was making AI-driven decisions that managers were unaware of?
Green flags and red flags — what good answers look like
A common trap: candidates answer with tool names instead of judgment. Saying "I use ChatGPT daily" tells you almost nothing. Explaining what they check in the output, what they'd never put in a prompt, and how they evaluate results in context — that's where real competency shows.
| Domain | Green flag | Red flag |
|---|---|---|
| Data privacy | "I anonymize scenarios before using any AI tool and only use company-approved platforms." | "I just use ChatGPT — it's public anyway." No awareness of data transfer or processing implications. |
| Bias and fairness | Names specific patterns they check for in job ads — gendered language, inflated credential requirements, exclusionary phrasing. | "AI is objective." No understanding that AI reflects training data biases. |
| Works council / governance | "I'd involve the works council before rollout and work toward a formal works agreement." | "That's a management decision, not HR's." Missing co-determination awareness. |
| Output quality | Describes a concrete review step: fact-check, tone evaluation, hallucination scan. | Treats AI output as directly usable without review. |
| Human decisions | "AI helps me draft. The hiring decision always sits with a person." | "If the AI scores the candidate highly, that's good enough." Treats AI as the decision-maker. |
The live scenario: 1 task, 10 minutes, real insight
Abstract questions can be rehearsed. Real competency surfaces under a small amount of pressure. A short live scenario reveals more than ten interview answers about how someone actually works with AI.
Task (written, 10 minutes):
- Topic: "Using an AI tool of your choice, write a 150-word interview invitation email for an HRBP role at a mid-size pharmaceutical company."
- Debrief (3 minutes): What did you input? What did you change in the output? What would you never have included in the prompt?
What the scenario reveals:
- Prompt design: Does the person write a structured, context-rich prompt — or a single sentence and then wonder why the output is generic?
- Output critique: Do they actively revise — tone, legal accuracy, inclusivity? Or accept the first draft?
- Data awareness: Do they mention unprompted what they wouldn't include — candidate names, internal salary ranges, confidential role context?
The same scenario works for recruiter roles (a job ad instead of an invitation), HR Ops (a survey result summary), or Head of People (a governance memo on a new AI tool).
The DACH regulatory framework: what candidates need to know
In Germany, Austria, and Switzerland, three regulatory frameworks shape AI use in HR. They overlap but regulate different aspects. Candidates at HRBP level and above should be able to navigate all three.
EU AI Act (high-risk classification)
AI systems used in employment decisions qualify as high-risk under Annex III of the EU AI Act. This includes CV screening, performance monitoring, and AI-assisted skills assessments. From August 2026, organizations must document how these systems work, how human oversight is maintained, and how affected individuals are informed. Crucially: the employing organization as deployer carries independent obligations — vendor compliance does not substitute for your own. For a practical overview of building the full training and governance stack, the AI Enablement in HR guide for DACH is a practical starting point.
GDPR
The GDPR runs alongside the EU AI Act. Applying AI to employee or candidate personal data typically requires a Data Protection Impact Assessment under Art. 35 GDPR. Data subjects retain access, correction, and deletion rights that AI processes cannot undermine. Entering personal data into external AI services — even via a conversational interface — may constitute an unlawful data transfer.
BetrVG (Germany)
Under § 87(1)(6) BetrVG, works councils have co-determination rights over any technical system capable of monitoring employee behavior or performance. The employer's intent is irrelevant — objective suitability for monitoring suffices. Since the 2021 BetrVG reform, under § 80(3) BetrVG, works councils can bring in an AI expert without a necessity review when AI is introduced. A formal works agreement (Betriebsvereinbarung) is in most cases the cleanest path to legal clarity.
Candidates who can explain these three layers — not as compliance jargon, but with concrete operational implications — demonstrate that they can introduce AI responsibly, not just use it.
How to score: a simple competency model
A 5-point scoring rubric helps objectify panel impressions and focus debrief conversations. It prevents articulate candidates without real substance from outscoring quieter experts with genuine practice.
| Score | Description | Typical behavior |
|---|---|---|
| 5 | Exemplary | Concrete examples, own lessons from mistakes, proactively names limits, knows the regulatory framework. |
| 4 | Strong | Good practical examples, solid risk and data privacy awareness. Minor gaps in compliance detail. |
| 3 | Adequate | Uses AI but reactively. Reviews outputs but without a system. Compliance knowledge is surface-level. |
| 2 | Development needed | Trusts AI outputs without review. Doesn't recognize data privacy risks. Works council not on radar. |
| 1 | Not suitable | Treats AI as the decision-maker. Would enter unprotected personal data into external tools. |
How to apply it: Score each competency domain separately. A score below 3 in data privacy or governance should be treated as a disqualifier for roles with AI rollout responsibility. A score below 2 in any domain warrants an immediate stop in the process — regardless of other strengths. Read the scores as a profile, not a sum. Someone scoring 5 on practical use and 2 on data privacy is not a good hire — it's a liability.
FAQ: AI interview questions for HR roles
What's the difference between AI knowledge and AI competency in an HR interview?
AI knowledge means knowing tools or being able to define terms. AI competency means understanding when AI helps, when it causes harm, and how to act responsibly in between. In an interview context, competency is what matters — and it can only be assessed through behavioral and scenario questions, not knowledge quizzes.
Should I ask about specific tools or about behavior?
Always behavior. Tool knowledge changes fast — whether it's ChatGPT, Copilot, or an industry-specific ATS, those tools may be irrelevant in a year. Behavior — reviewing outputs, questioning results, setting boundaries — stays relevant. Ask: "What did you change before you used the AI output?" That tells you more than "Which AI tools do you know?"
Does the EU AI Act apply to small and medium-sized companies?
Yes. High-risk obligations under the EU AI Act apply to all deployers regardless of company size. SMEs face lower penalty caps, but the compliance requirements themselves — human oversight, documentation, transparency — apply equally. For HR candidates, this means the same obligations apply to a mid-size regional company as to a large corporate.
How do I tell if a candidate is just using buzzwords?
Ask for a concrete example: "Can you describe a situation where something went wrong?" Or: "Tell me about a time you rejected an AI output." Candidates with real experience can answer immediately. Those who've only read about it start to struggle. The follow-up question "What did you actually change?" is the most reliable filter.
How many AI questions should an HR interview include?
It depends on the role. For recruiters, 3–4 targeted behavioral questions plus a short scenario is enough. For HRBPs, 5–6 questions make sense, including at least one governance question. For Heads of People, 6–8 questions with a strategy and regulatory focus. AI should be treated like any other core competency — not tacked on as an afterthought at the end of the interview.
What if the interview panel has limited AI experience?
Use observable thresholds rather than comparative judgments. A panel without direct AI experience can still assess: Did the person name data privacy risks? Did they mention works council obligations? Did they revise the live scenario output or accept it unchanged? These are behavioral signals that don't require AI expertise in the panel to evaluate.
Conclusion
AI competency in HR isn't about knowing tools — it's about judgment. Organizations hiring for HR roles need candidates who use AI safely, recognize its limits, and take accountability for data privacy, fairness, and governance. That's testable with targeted behavioral questions and a short live scenario. Structured scoring turns impressions into defensible decisions.
Jürgen Ulbrich
Jürgen Ulbrich has more than a decade of experience in developing and leading high-performing teams and companies. As an expert in employee referral programs as well as feedback and performance processes, Jürgen has helped over 100 organizations optimize their talent acquisition and development strategies.
