GPT has killed competency-based interviewing

By Paul Marsh (not AI)

For years, competency-based interviewing has been the gold standard of recruitment. Candidates are asked to describe a time when they demonstrated leadership, resilience, influencing skills or problem solving. It's all based on the premise that past behaviour predicts future performance.

But the rise of AI has fundamentally changed the way candidates prepare for interviews. Anyone can ask ChatGPT to generate compelling examples for virtually every competency in existence. Candidates can rehearse polished answers, memorise perfect structures such as STAR, and deliver stories that sound authentic even if they are heavily enhanced or partly fictional. The interview has become less of a test of capability and more of a test of preparation.

People also naturally present their successes while conveniently forgetting the context, the mistakes and the support they received from others. Traditional competency interviews reward storytelling ability as much as genuine competence. If organisations want to identify leaders who can actually perform, they need to move beyond rehearsed examples and test how people think in real time. Here are three alternatives:

The first element is live scenario questioning.

Rather than asking what a candidate did three years ago, present them with a realistic business challenge they could face in the role today. Ask them to work through it aloud, give them incomplete information, introduce changing circumstances. Observe not only the conclusion they reach but the quality of their thinking, prioritisation, communication and judgement.

Strong leaders rarely have perfect answers immediately. What distinguishes them is how they process uncertainty, balance competing priorities and adapt when new information emerges. Live scenarios reveal these behaviours far more effectively than rehearsed stories ever can.

The second element is failure questioning.

Most interviews spend far too much time celebrating success. Yet leadership is defined just as much by failure. Ask candidates about decisions that went wrong, appointments they regret, projects that failed or conflicts they mishandled. Then explore what happened next. Did they accept responsibility or shift blame? What did they learn? What changed in their leadership style afterwards? Can they demonstrate genuine self awareness?

The ability to reflect honestly on failure is one of the strongest indicators of emotional maturity, resilience and future growth. These are qualities that polished competency examples rarely expose.

The third element is constructive challenge.

Too many interviews become comfortable conversations. Candidates deliver prepared answers and interviewers move politely to the next question. Instead, interviewers should respectfully challenge assumptions, probe inconsistencies and ask candidates to defend their reasoning.

"You said you empower others, yet this example suggests you made every key decision yourself. Help me understand that."

"What other options did you consider?"

Challenge is not about catching candidates out. It's about discovering how they respond when their thinking is questioned. Senior executives spend much of their working lives defending decisions, influencing stakeholders and adapting under pressure. Interviews should reflect that reality.

Competency based interviewing is not completely obsolete. Past experiences still provide valuable context. However, relying on historical examples alone is no longer enough in an era where candidates can arrive armed with AI generated content and perfectly rehearsed answers. The organisations that hire the best leaders will increasingly focus on evidence rather than storytelling. They will combine past examples with live problem solving, honest exploration of failure and thoughtful challenge. They will assess not only what candidates claim they have done, but how they think, how they learn and how they respond when the script disappears.