The European AI Act Is Now Law: What Changes for Companies from August 2, 2026
From August 2, 2026, the most binding obligations of the European AI Act come into full effect. Transparency, mandatory training, and new corporate responsibilities: here is what businesses and workers need to know.
There is a date circled in red on the calendars of compliance officers across Europe: August 2, 2026. From that day forward, the European Union's Artificial Intelligence Regulation — better known as the AI Act — enters its most concrete and enforceable phase. This is not one of those rules that linger on paper for years, waiting for implementing decrees that never materialize. This time, inspections will happen, penalties exist, and companies that have not yet acted risk finding themselves completely unprepared at a moment when artificial intelligence has already quietly embedded itself into daily work processes.
But what actually changes? The core aspects are essentially two, and they concern any organization using tools such as ChatGPT, Microsoft Copilot, Google Gemini, or any other large language model-based system.
The first is transparency. European regulation makes it clear that users must be informed when they are interacting with an artificial intelligence system rather than a human being. This means that if a company uses a chatbot or virtual agent to respond to customers on its website, it must explicitly state that the entity responding is not a person but an algorithm. Knowing this internally is not enough: the obligation runs toward the end user. The same principle extends to AI-generated content — images, videos, documents, presentations — which must be clearly labeled as such. Passing off machine-produced material as human-created work becomes, in every practical sense, a regulatory violation.
The second point is perhaps the one that surprises people most, even those already familiar with the broad outlines of the AI Act: mandatory training. The obligation, in fact, came into effect as early as February 2025, but it is with the start of inspections scheduled for August 2026 that the matter becomes genuinely urgent. Companies — not just technology firms, not just software developers, but every organization that uses AI-based tools — are required to ensure that their employees maintain an adequate level of digital literacy regarding artificial intelligence.
This does not mean a single standardized course for everyone. The regulation does not prescribe a uniform program to be followed collectively: training must be specific, proportionate to each employee's role, and calibrated to how they actually interact with AI systems. Someone who uses a conversational assistant to draft emails has different training needs from someone who relies on automated systems for recruitment decisions or the analysis of sensitive data. Artificial intelligence, in short, stops being a technological accessory and becomes a professional competency subject to organizational accountability.
This is the first time in the history of European legislation that understanding how a technology works is no longer merely a useful skill but a legal obligation. And this profoundly changes how companies must approach AI adoption: not as a competitive advantage to be seized quickly, but as a responsibility to be managed with care, awareness, and structure.
The question, then, is simple and direct: in your company, has anyone already explained how to use artificial intelligence correctly, safely, and responsibly? If the answer is no, the window for action is narrowing fast.
Source: Regulation (EU) 2024/1689 of the European Parliament and of the Council on Artificial Intelligence (AI Act), published in the Official Journal of the European Union.