Artificial intelligence in companies: innovation that needs management
Artificial intelligence in companies: innovation that needs management
The year 2026 has definitively ended the debate about whether the regulation of artificial intelligence will affect companies. The AI Act is already partially effective, and the responsibility for the use of artificial intelligence now lies directly with the management of organizations. At the same time, many companies use AI in a way that, on closer inspection, would not stand up to an internal audit – let alone a regulator. The question is therefore not whether to deal with AI, but whether you really have it under control.
Artificial intelligence has become a common part of the daily operation of companies in recent years. In 2026, organizations use it not only to automate processes or analyze data, but also to evaluate employees, manage customer relationships, set prices, and support strategic decision-making. However, it is this "invisible" integration of AI into key processes that poses one of the biggest risks today.
The AI Act is no longer just a planned regulation. Selected provisions of the Constitution are effective and others are being applied gradually. Companies that use AI must be able to demonstrate that they know where and how they use AI, what risks are involved, and who is responsible for these decisions. For many organizations, this finding is uncomfortable – especially when they realize they don't have the answer to these questions.
From practice, we can see cases today where the HR department uses AI-based tools to pre-select candidates, without it being clear on the basis of which criteria the system makes decisions and whether it does not exhibit discriminatory behavior. In other companies, AI tools are used to evaluate employee performance or to propose personnel measures, while the responsibility for the final decision remains formally with the manager, who, however, does not understand the functioning of the algorithm and does not have the opportunity to realistically verify its outputs.
Another common scenario is the use of AI in customer communication. Today, automated systems decide what offers the customer will receive, what prices will be offered to them or how their complaints will be resolved. If such a system misjudges the situation, the impact is not on the technology supplier, but on the responsibility of the organization itself. In addition, in the context of the AI Act, it may be a high-risk system for which regulation requires an increased level of control and documentation.
We also encounter cases where employees use generative AI tools to prepare analyses, presentations or materials for management decision-making, without there being clear rules for working with data. This can allow sensitive or commercially significant information to inadvertently get out of the organization. From the point of view of regulation and the company's reputation, this is a risk that management often realizes only when it is too late.
The common denominator of these situations is not bad intentions, but lack of control. Artificial intelligence entered processes faster than rules, responsibilities and control mechanisms were created. However, the AI Act clearly states in 2026 that this approach is no longer acceptable.
Companies that want to use AI in the long term and safely must move from random use to systematic management. This means having an overview of all AI systems in the organization, understanding their purpose and risk, having clearly defined responsibilities, and being able to demonstrate compliance with applicable regulations. This is not about limiting innovation, but about creating an environment in which innovation can operate without unnecessary legal and reputational risks.
In this area, BDO builds on its long-term experience in regulation, risk management and information systems. We help organizations translate the requirements of the AI Act into practically functioning processes. We focus on ensuring that company management has a real overview of how AI is used, who is responsible for it and how the associated risks are managed. Our support also includes aligning the requirements of the AI Act with other proven frameworks, such as the NIST AI Risk Management Framework or ISO/IEC 42001.
The year 2026 is a turning point for many companies. Artificial intelligence has definitely become an area that cannot be left only to IT or individual departments. It is a topic of strategic management. Organizations that grasp it in a timely and systematic manner will gain a stable basis for safe innovation. The others risk encountering AI Act for the first time only during a review or when solving a specific problem.
BDO is ready to be a partner who will help you grasp this area with the necessary perspective, expertise and in accordance with the regulations already in force. We will be happy to discuss with you without obligation how the use of artificial intelligence specifically applies to your organization and which procedure makes the most sense in 2026.