AI Policy and Responsible Use of Artificial Intelligence

Last updated: 15/04/2026

This AI Policy explains how we use artificial intelligence (AI) within the Platform, which privacy safeguards we apply, and which rules apply to organisations, HR teams and managers who work with the results.

Contents

1. Short summary

In 1 minute

We use artificial intelligence (AI) to provide users of the Platform with additional feedback on their talents and development points. This AI is supportive and does not make autonomous decisions about individuals. We aim to reduce privacy risks by pseudonymising identifying data as much as possible, for example by replacing names with numbers or internal codes. In addition, we use a professional paid AI service in which submitted data is not used to train or improve models. AI feedback may never be the sole basis for decisions such as recruitment, evaluation, promotion, task allocation or dismissal.

Detailed

This AI Policy explains how we use AI within the Platform, which safeguards we take regarding privacy and data protection, and which rules apply to organisations that use the Platform’s results.

The Platform enables users to take a talent assessment, receive talent insights, participate in teams and receive additional AI-supported feedback. The AI functionality is intended to make insights clearer, more accessible and more concrete. It is not intended to classify people for disciplinary, legal or employment-related purposes.

In line with the AI Act, we apply the following core principles:

  • transparency: users must know when AI is being used;
  • human oversight: AI does not replace human judgment;
  • data minimisation: only necessary data is processed;
  • privacy by design: where possible, names are replaced by numbers or other identifiers;
  • limited use: AI feedback is supportive and may not be used as the sole basis for impactful decisions.

Back to contents

2. How we use AI on the Platform

In 1 minute

AI is used on the Platform to formulate additional feedback on talents, test results and development opportunities. AI therefore helps clarify or structure information, but it does not decide on someone’s career, employment or position within an organisation. AI output is advisory and supportive. There is no fully automated decision-making with legal or similarly significant effects based solely on this AI output.

Detailed

Within the Platform, AI is used to generate textual feedback based on talent insights, test outcomes and related input. This feedback may, for example:

  • summarise talents;
  • explain strengths;
  • suggest development areas;
  • structure descriptive feedback in clear language.

The AI output always has a supportive character. This means, among other things:

  • the AI output is not binding;
  • the AI output is not final or absolute;
  • the AI output must always be read in context;
  • the AI output may not automatically be turned into a decision affecting someone directly.

Where relevant, a human user must interpret the AI output with appropriate caution. This is especially true in a professional context in which results may be viewed by HR, managers or other organisational stakeholders.

In terms of privacy law, processing test results to generate talent profiles or insights may under certain circumstances be considered profiling. That is why we avoid any situation in which AI outcomes, without human involvement, lead to decisions with legal or similarly significant effects.

Back to contents

3. Privacy protection: pseudonymisation and no training on personal data

In 1 minute

We take privacy protection seriously. Wherever possible, we replace names and directly identifying data with numbers, codes or other internal identifiers before data is used within AI processes. In addition, we use a professional AI service in which submitted data is not used to train or improve models. This significantly reduces privacy risks.

Detailed

To protect user privacy, we apply pseudonymisation wherever possible. This means that identifiable information, such as names, is replaced as much as possible by numbers, codes or other internal references. This reduces the risk of direct identification.

Pseudonymisation is an important measure, but it is not the same as full anonymisation. Where data can still reasonably be linked back to an individual, it remains personal data under the GDPR. We therefore continue to handle such information with the required care and security.

For our AI functionality, we use a professional paid AI solution. Contractually and operationally, submitted data is not used to train or improve AI models. This is an important safeguard for users and client organisations, because personal data or pseudonymised data provided to the service does not enter a general model training stream.

In addition, we apply the principle of data minimisation:

  • we limit input to what is necessary;
  • where possible, we avoid unnecessary identifying information;
  • we apply appropriate technical and organisational security measures.

Back to contents

4. Transparency towards users

In 1 minute

When AI is used to generate feedback, users must be clearly informed. That is why we make it visible when certain feedback or explanations are AI-supported. Users must understand when AI plays a role and what its limitations are.

Detailed

Transparency is a core principle of responsible AI use. Users should be able to understand in clear language:

  • that AI is being used;
  • what exactly AI is used for;
  • that the AI output is advisory;
  • that the AI does not make autonomous decisions;
  • that their personal data is not used to train or improve AI models;
  • that identifying data is pseudonymised where possible.

For that reason, we may display short notices, info icons or explanatory texts within the Platform when AI functionalities are used. Such notices clarify that a user is dealing with AI-supported output and help avoid misunderstandings.

Back to contents

5. What AI on this Platform explicitly does not do

In 1 minute

Our AI is not intended to assess people’s value as employees, detect emotions in the workplace, or automatically decide on dismissal, promotion or other HR measures. The Platform is not designed for prohibited AI practices and is not intended for manipulative or discriminatory uses.

Detailed

The Platform and the AI functionalities offered on it are not intended for:

  • making fully automated HR decisions;
  • detecting or inferring emotions in the workplace;
  • assessing the reliability, loyalty or “suitability” of employees outside the context of talent development;
  • exercising undue pressure or manipulation on users;
  • creating social ranking or similar scoring mechanisms;
  • using AI feedback as the sole basis for measures with significant consequences.

The AI Act treats employment and worker management as a particularly sensitive area. For that reason, the output of this Platform may not be repurposed for aims for which it was not designed.

Back to contents

6. Guidelines for organisations, HR and managers

In 1 minute

Organisations may use the Platform’s insights for development, guidance, teamwork and dialogue. They may not use these insights as the sole or decisive basis for decisions such as dismissal, demotion, promotion, sanctions or rejection of candidates. In a workplace context, AI is especially sensitive under the AI Act.

Detailed

The Platform is designed as a tool for talent development, self-insight and team support. Organisations, HR teams and managers may only use the results and AI feedback in a way that is compatible with that purpose.

Appropriate use includes, for example:

  • development conversations;
  • coaching;
  • career reflection;
  • understanding complementary talents within a team;
  • supporting collaboration and communication;
  • voluntary reflection on strengths and areas for improvement.

Inappropriate or prohibited use includes, for example:

  • basing dismissal decisions exclusively or mainly on AI feedback;
  • making promotion or non-promotion dependent on automated talent labels;
  • enforcing task allocation on the basis of AI output without human assessment;
  • linking disciplinary measures to test or AI results;
  • automatically excluding candidates or employees based on talent output;
  • treating AI output as objective truth without context or possibility of challenge.

Back to contents

7. What HR and managers are allowed to do

In 1 minute

HR and managers may use the Platform’s output as a conversation starter and as a tool for development. They may use talent insights to better understand teams, support collaboration and guide employees more effectively. The condition is that results are never applied blindly or mechanically, and that human context always comes first.

Detailed

HR and managers may use the Platform’s results for legitimate, proportionate and development-oriented purposes, such as:

  • preparing a development or feedback conversation;
  • discussing talents and work preferences with the employee;
  • strengthening team dynamics and collaboration;
  • making complementary strengths visible within a team;
  • supporting learning and growth trajectories;
  • facilitating dialogue on division of work, communication or mutual alignment.

The following conditions always apply:

  • the employee must understand that these are supportive insights;
  • the output must be read together with human observation and context;
  • an employee must have room for nuance, correction or challenge;
  • the organisation must handle confidentiality and access rights carefully;
  • the use must fit within the privacy notice, internal HR processes and applicable law.

Back to contents

8. What HR and managers are not allowed to do

In 1 minute

HR and managers may not use the Platform to automatically assess, exclude or sanction anyone. Talent insights and AI feedback may not serve as the sole basis for dismissal, negative evaluations, promotion decisions, recruitment rejection or major changes in job content.

Detailed

Organisations, HR and managers may in any event not:

  • dismiss an employee based solely on AI feedback or talent profiles;
  • automatically refuse promotion or salary increase based on test or AI results;
  • treat the output as an objective or medical judgment of someone’s suitability;
  • rank or label employees in a discriminatory, stigmatising or disproportionate way;
  • use talent feedback as a hidden selection tool without transparency;
  • draw conclusions from AI output about health, mental state or emotional stability;
  • attempt to detect or infer emotions in the workplace through AI;
  • force employees to accept AI insights as truth without room for discussion or nuance.

In practice, this also means that the following uses are not allowed or not appropriate:

  • “This employee scores low on leadership, so no promotion.”
  • “The AI feedback shows doubt, so we end the employment relationship.”
  • “This candidate does not fit the culture according to the talent profile, so reject.”
  • “The team output shows insufficient suitability, so reassign without discussion.”

Back to contents

9. Mandatory caution in an HR context

In 1 minute

Anyone using AI in an HR or workplace context must act with particular caution. This means being transparent, maintaining human oversight, respecting privacy, avoiding discrimination and assessing whether a given use falls under stricter AI Act obligations.

Detailed

The AI Act identifies AI systems used in areas such as employment, worker management and access to self-employment as particularly sensitive. Depending on how an organisation uses an AI system in practice, stricter obligations may apply.

For organisations, this means at minimum that they must:

  • have a clear purpose for the use of AI;
  • only use AI output in a proportionate manner;
  • ensure meaningful human involvement;
  • provide sufficient information to employees;
  • internally define who has access to which data;
  • assess whether the use of AI amounts to a high-risk deployment;
  • comply with privacy, employment and equality rules.

Even where the Platform itself is intended as a supportive development tool, improper use by the organisation may still create legal risks.

Back to contents

10. Practical rules of conduct for organisations

In 1 minute

Use AI output only as a tool, never as the final decision-maker. Limit access to those who genuinely need the information. Discuss results with the employee concerned. Document internally what the Platform may and may not be used for. And never use AI for dismissal, sanctions or automatic exclusion.

Detailed

  1. Use the Platform only for development and team support.
    Not for discipline, control or automated selection.
  2. Ensure human review.
    Any interpretation of talent insights or AI feedback must be made by a human and must take into account context, job content and the possibility of challenge.
  3. Never use AI output as the sole basis for a decision.
    This applies in particular to dismissal, promotion, evaluation, task allocation, remuneration and recruitment.
  4. Strictly limit access to authorised persons.
    Not everyone in the organisation needs the same level of access.
  5. Respect transparency towards employees.
    Employees must understand what is visible, who can see it and for what purpose it is used.
  6. Avoid repurposing beyond the original purpose.
    What is intended for talent development may not quietly be turned into a sanction or selection tool.
  7. Set internal governance rules.
    Define in writing which HR or management roles may use the Platform and for what purpose.
  8. Escalate when in doubt.
    Where the intended use concerns sensitive HR decisions, legal and privacy advice should be obtained first.

Back to contents

11. Final statement

In 1 minute

Our AI is designed to support people, not to decide about them. That is why we combine transparency, privacy safeguards, pseudonymisation and the principle that personal data is not used to train or improve AI models. For organisations, there is a clear boundary: use talent insights to strengthen people, not to automatically assess or sanction them.

Detailed

We believe AI can only be used responsibly when:

  • users understand that AI is being used;
  • privacy is genuinely protected;
  • personal data is not used for model training;
  • human dignity and context remain central;
  • organisations do not misuse AI for one-sided or harmful HR decisions.

For that reason, the following principle applies on our Platform:

The results and AI feedback provided by the Platform are supportive in nature and may not be used as the sole basis for decisions regarding recruitment, evaluation, promotion, task allocation or termination of the employment relationship. Personal data processed in connection with this AI functionality is furthermore not used to train or improve AI training models.