Assessments play a major role in the selection and development of employees. Clients expect an objective, consistent evaluation; candidates want equal and fair treatment. With the massive deployment of AI, now also within the HR sector, this calls for demonstrable transparency and human oversight.

Also read: ilionx scores high in 2026 IT Sourcing Study Netherlands

Standardization of role plays and scoring

ChatLab automates and optimizes components that are traditionally highly protocolled and time-consuming. Think of role plays to simulate practical situations and structured interviews for competency measurements. By having an AI avatar conduct the conversation with the candidate via a voice-to-voice system, a fixed, reliable format is used based on pre-established criteria. The AI assessment can also be conducted remotely, making the processes faster and more flexible.

The core of the profession does not change. The psychologist remains ultimately responsible for the interpretation and explanation of the scores, conducts the in-depth interview with the candidate, and writes the advice. The final evaluation remains based on professional judgment, only with more consistency and room for depth.

'Decisions about someone's career should not be left to AI,' says Krans. 'Our approach has been based on scientific validation and human professional judgment for a hundred years. With this, we integrate that same care into the AI environment.'

In this way, employers can make better-informed choices in the selection process and candidates can start in the role that fits their talent.

Also read: Huxe AI: interactive podcasts from email and agenda

Preventing and monitoring AI bias

To prevent AI bias within ChatLab, the agency actively monitors the models for deviations and possible distortions. 'Bias is not a one-time check, but a continuous process,' says Richard Osinga, Head of Innovation at LTP Business Psychologists. 'Because we work with explicit criteria and transparent parameters, we can adjust the model in a timely manner.'

ChatLab has been tested more than five hundred times in simulations and in four pilots with three launching customers. Scores have been validated based on the more than 100,000 anonymous profiles in its own platform. The AI application within ChatLab is set up within the frameworks of the GDPR and the European AI Act, thus meeting the requirements in the areas of transparency, human oversight, and risk management. External agencies have conducted technical and legal due diligence.

Also read: Odido hacked: 6.5M people – this is the real risk