There is a conversation happening in many organisations. Someone in HR, IT or Sustainability identifies the need to train employees in the responsible use of AI. They prepare the proposal. Then the executive team asks the inevitable question:
Why should we invest in this now?
The answer often focuses on digital culture, competitiveness or keeping up with technological change. Those are valid arguments. But they are not the strongest ones because the real question leadership is asking is not whether AI training is valuable, It is whether it is necessary now.
And there is a far stronger answer than any competitiveness argument: the law already requires it.
Article 4 of the AI Act: The Obligation Few Organisations Are Talking About
Article 4 of Regulation (EU) 2024/1689 requires organisations deploying AI systems to take measures to ensure that their staff possess a sufficient level of AI literacy for their roles and responsibilities.
This provision has applied since 2 February 2025. It was not affected by the Digital Omnibus package or any subsequent proposal to postpone implementation.
In practical terms, this means that if your organisation uses AI systems—which, by 2026, applies to almost every organisation—it has a legal obligation to ensure that the people using those systems receive appropriate AI literacy training.
This is not a recommendation or a best practice, It is a legal requirement.
Since August 2025, compliance with the AI Act has been supervised in Spain by the Spanish Agency for the Supervision of Artificial Intelligence (AESIA). It is worth presenting this argument carefully, especially to a well-advised executive team.
The real strength of Article 4 is not that organisations are automatically fined for failing to train employees. Rather, if an AI-related incident occurs, the absence of appropriate training can significantly weaken the organisation’s position as a deployer. Training is not simply a compliance exercise, It is evidence of due diligence.
And when something goes wrong, the difference between having trained your people and not having done so may determine whether the incident is viewed as responsibly managed—or as evidence of organisational negligence.
The Risk Argument Every Executive Team Understands
Executive teams manage risk. That is why the most persuasive argument is not educational, It is operational, financial and reputational. Responsible AI training directly reduces three categories of organisational risk.
Regulatory Risk
Failure to comply with the obligations imposed on deployers under the AI Act may lead to administrative penalties of up to 3% of worldwide annual turnover.
For an organisation with annual revenues of €10 million, that could represent penalties of up to €300,000. Training represents a fixed and predictable investment. Regulatory sanctions do not.
Operational Risk
An employee using AI without appropriate judgement may trigger a personal data breach, introduce discriminatory outcomes into decision-making, or unintentionally disclose confidential information.
Any one of these incidents is likely to cost significantly more than preventive training.
Reputational Risk
Customers, partners and investors increasingly pay attention to how organisations use AI.
A public AI-related incident can cause long-term reputational damage. Training transforms employees from a potential source of risk into an active line of defence. In other words, responsible AI training is one of the most cost-effective preventive measures an organisation can take against risks that already exist.
The Mistake of Presenting It as Tool Training
One reason these proposals often fail is because they are presented simply as AI training:
“Learn better prompts.”
“Discover the latest AI tools.”
That message rarely convinces executive teams because it sounds like individual productivity training rather than organisational risk management.
Responsible AI training is something different. It does not focus on teaching tools, It develops the organisational judgement needed to decide when AI should be used, what data may be processed, what level of human oversight is required, and how the organisation should respond when something goes wrong.
That is not an employee benefit, It is organisational infrastructure.
How to Present the Proposal to Leadership
Emotional arguments about digital transformation are rarely enough.
What works is a structured risk assessment.
Identify the organisation’s exposure scenarios, estimate the potential impact of each, compare those risks with the cost of the training designed to reduce them.
Then support the proposal with the regulatory evidence. Reference Article 4 of the AI Act and its date of application. Explain the supervisory role of AESIA. If the organisation reports against ESG frameworks, demonstrate how responsible AI governance contributes to documented governance practices and organisational accountability.
Finally, present a concrete programme with a tangible outcome. Not simply AI training as an abstract concept.
A structured programme with defined objectives, a clear duration and measurable deliverables.
One of the strongest deliverables is an organisational AI Use Policy formally approved by senior leadership. That transforms training from a one-off learning activity into a lasting organisational asset.
Frequently Asked Questions
Is AI training mandatory for employees in Spain?
Yes. Article 4 of the AI Act requires organisations deploying AI systems to ensure that staff possess an appropriate level of AI literacy for their responsibilities.
The obligation has applied since 2 February 2025 and applies regardless of the organisation’s size or sector.
What happens if my organisation does not provide AI training?
The obligation under Article 4 is already in force, and compliance with the AI Act is supervised by AESIA in Spain.
Beyond the possibility of regulatory penalties, the most immediate consequence is that the absence of training may significantly increase the organisation’s liability if an AI-related incident is investigated or challenged.
Administrative fines for deployers may reach 3% of worldwide annual turnover, depending on the circumstances.
What is the cost of doing nothing?
The cost of an unmanaged AI incident may include regulatory penalties, reputational damage, loss of customers and legal expenses.
Preventive AI training represents a fixed, predictable investment that is consistently lower than the cost of any one of those risks occurring individually.
