The AI Act and Entrepreneurs: Why It Is Not Enough to Use Artificial Intelligence Without Rules

16. 06. 2026

Artificial intelligence has become a common working tool in business. Companies use it in marketing, text creation, customer support, data analysis, employee recruitment, internal administration, and the automation of decision-making processes. This is precisely why it is important to monitor the European regulation on artificial intelligence, known as the AI Act, which is gradually introducing rules for developers, providers, and users of artificial intelligence systems.

The AI Act is based on a risk-based approach. Not every use of AI will be regulated in the same way. Many common tools will fall into the category of minimal or low risk, while other systems may be subject to specific requirements, especially if they affect individual rights, safety, employment, education, critical infrastructure, or other sensitive areas. For entrepreneurs, it will therefore be essential first to understand what AI is actually being used for within the company.

From a practical perspective, entrepreneurs will in the near future primarily encounter transparency requirements. The AI Act envisages that, in certain cases, a person must be informed that they are communicating with a machine rather than a human being. In the case of generative AI, the labelling of content created by artificial intelligence is also addressed, particularly where deepfake content or texts intended to inform the public about matters of public interest are concerned. These rules are relevant not only for technology companies, but also for marketing agencies, e-shops, media, influencers, and companies using AI in customer communication.

It is also important to distinguish official corporate use of AI from so-called uncontrolled use by employees. If employees enter business data, personal data, contracts, internal documents, or client information into publicly available AI tools, this may create problems not only from the perspective of the AI Act, but also from the perspective of personal data protection, trade secrets, confidentiality, and contractual obligations towards business partners.

Entrepreneurs should therefore not wait until a specific obligation begins to be directly enforced. A practical first step is to create an internal AI use policy. Such a policy should determine which tools may be used within the company, what data employees must not enter into them, who is responsible for checking outputs, when content must be labelled as created or modified by AI, and how the company will proceed in sensitive areas, such as employee recruitment or customer assessment.

Special attention should also be paid to responsibility for outputs. The fact that a text, image, analysis, or recommendation was created by artificial intelligence does not mean that the company bears no responsibility for it. If an entrepreneur uses an AI output in advertising, a commercial offer, contractual documentation, or communication with a customer, they should review it in the same way as any other corporate output. AI may be a tool, but not an automatic excuse for inaccuracy, misleading information, or infringement of third-party rights.

In conclusion, entrepreneurs should not view the AI Act as a distant regulation intended only for large technology companies. Even ordinary use of AI within a company already raises questions of transparency, data protection, liability, copyright, and internal control. Companies that set rules in time can use AI more effectively and safely. Companies that allow AI to be used in a completely uncontrolled manner, on the other hand, risk turning a useful tool into a legal and reputational problem.

All articles