How to use AI safely
at work – training for employees

The training is available in on-site or online format – depending on your preferences. It is conducted in an interactive formula that provides space for discussion and questions. The programme has been designed to make the training as effective and concise as possible – the whole session takes up to 2 hours.

Each participant will receive a certificate confirming their participation in the training, which shall include a detailed scope of training, date of training and information on the trainer.

Yes, pursuant to Article 4 of the AI Act, companies that create or use artificial intelligence systems must ensure their employees have the appropriate knowledge and skills. This obligation also applies to persons cooperating with such a company.

The law does not impose a specific form for acquiring these competencies, but training is the simplest and most practical way to fulfil this obligation. It enables employees to consciously use AI, better control it, and avoid mistakes. Lack of adequate preparation can lead to violations of regulations or serious problems related to improper use of AI.

The regulations do not specify how often training should be conducted, but it should take place regularly and be continually adapted to technological and legal changes. Good practice is to organise cyclical training – e.g. once a year – as well as additional sessions in situations such as implementing new systems or a change in an employee's position. Evaluating training effectiveness is also key, as what counts is not only participation, but above all real results. Such an approach helps the company maintain a high level of readiness and minimises the risk of errors related to new technologies.

AI training should cover all persons who have contact with AI systems. This applies to creators and operators of these systems, end users, and external service providers. It is particularly important to train employees in HR, IT, legal, and compliance departments, as well as decision-makers responsible for implementing and supervising AI use. It is also important for people using AI outputs, such as analysts or department managers, to understand how these systems work. Even people indirectly affected by AI should have basic knowledge to understand the consequences of AI-made decisions.

AI safety training should explain in an accessible way how AI systems work and how to use them effectively and responsibly in practice. Key topics should include: interpretation of AI-generated results, identification and assessment of risks, and accountability for decisions taken. Discussion of ethical and legal aspects is also important, in line with Art. 3(56) AI Act, which emphasises the importance of knowledge, skills, and awareness related to the benefits, risks, and possible harms of AI use.

The training should also teach how to identify high-risk situations and how to respond to errors and irregularities. To increase practical value, it is worth using real examples and scenarios from the relevant industry. The scope of training should be adapted to participants' roles, technical knowledge, and the specifics of their industry.

Uncontrolled use of AI in a company carries significant legal, operational, and reputational risks. Lack of oversight can lead to violations of regulations such as GDPR or the AI Act, which threatens high penalties, as well as civil and administrative liability. Operational threats include erroneous AI decisions, lack of transparency, risk of discrimination, and non-compliance with regulations. Such situations can lead to financial losses, loss of trust, and reputational damage. Particularly risky is the use of AI in areas requiring special responsibility – e.g. recruitment, healthcare, or insurance – without human oversight. A responsible approach to AI use should therefore be a permanent element of risk management in every company.

Yes, the employer is responsible for the way in which employees and external entities use AI-based tools on behalf of the company. Even if an employee makes an unintentional mistake, the lack of training, clear policies, or adequate supervision may be considered negligence on the part of the employer. It is therefore important for the company to implement appropriate policies, train employees, and monitor AI use to ensure compliance and reduce risk.

The employer is required to supervise employees' use of AI tools. They should know which systems are used in the company, who uses them, for what purpose, and what the consequences may be. It is important to maintain a register of tools used, assign responsibility for them, and regularly assess the associated risk. Control and monitoring mechanisms, incident reporting, and internal audits are also necessary. Supervision should be adapted to the specifics of the company – its industry, technology, and employees' tasks. In practice, the privacy by design principle should be applied, ensuring data protection at the stage of designing and implementing AI systems. All AI systems should also be regularly reviewed and updated, with complete and compliant documentation.

It is worth implementing clear and effective internal procedures to ensure safe and lawful AI use. The foundation should be clearly defined principles for AI use and an obligation to document its applications: where, when, and for what purpose the technology is used. Maintaining a register of all AI systems used is also a good solution.

Procedures for incidents, mechanisms for reporting irregularities, and regular compliance checks are necessary. In everyday practice, implementation checklists, compliance tests, and the "human-in-the-loop" principle – requiring human approval of key AI decisions – prove effective.

Equally important is ongoing monitoring of AI systems, developing employees' competencies, and incorporating AI topics into the risk management strategy. This allows the company to better control AI use and effectively respond to emerging challenges.

AI training helps build a culture of responsible use of new technologies in the company. It increases employees' awareness, develops their digital skills, and teaches a critical approach to AI-based tools. This makes it easier to avoid mistakes arising from ignorance and to use technologies ethically, in accordance with regulations and the company's values. Such an attitude fosters trust – both within the team and in relations with clients, business partners, and supervisory authorities.

AI training should be a permanent element of the compliance and risk management system, as AI has a real impact on decisions that may carry legal or social consequences. Knowledge of the principles of safe and responsible use of these technologies is today not only good practice but also a regulatory obligation under the AI Act. Appropriate training helps reduce legal, operational, and reputational risks, supports regulatory compliance, and reinforces the company's ethical and transparent approach to new technologies. It is not only a safeguard, but also an element of building competitive advantage and stakeholder trust.