Artificial intelligence is already part of the daily work of many companies. However, its use requires clear and responsible criteria. For this reason, the European Union approved the European Artificial Intelligence Regulation, known as the AI Act, which establishes obligations for companies that develop or use certain AI systems.
From 2 August 2026, new transparency obligations will start to apply. Their objective is to ensure that people know when they are interacting with AI or when content has been generated or manipulated using this technology.
The European Commission has published new guidelines to facilitate their application. However, the specific obligations will depend on the type of tool and on the role each company plays in its development or use.
Which obligations start to apply on 2 August?
The new transparency rules focus on certain interactive and generative systems.
Among other aspects, they establish that people must be informed when they interact directly with artificial intelligence, for example through certain virtual assistants or chatbots.
In addition, providers must incorporate mechanisms that allow certain AI-generated or AI-manipulated content to be automatically identified.
Companies that use these systems must also inform people when they present:
- Artificially manipulated images, audio or videos, known as deepfakes.
- AI-generated texts on matters of public interest when they have not been reviewed by a person.
- Emotion recognition or biometric categorisation tools, when their use is legally permitted.
This does not mean that any use of artificial intelligence automatically requires a warning. It will be necessary to analyse which tool is being used, for what purpose and who interacts with it.
How can the use of AI affect the workplace?
Companies are increasingly incorporating artificial intelligence tools into processes related to recruitment, internal communication and people management.
For example, they may use assistants to respond to candidate queries, internal chatbots for employees or systems that help prepare communications and employment-related documentation.
For this reason, companies should review whether these tools interact directly with people and whether they clearly inform them that artificial intelligence is being used.
It is also advisable to identify which applications are being used in each department. In many organisations, some tools may have been incorporated without a common internal procedure or without a prior review of their employment and data protection implications.
What changes with the Digital Omnibus Regulation on AI?
The Digital Omnibus Regulation on AI entered into force on 27 July 2026 and modifies several aspects of the timetable and obligations initially provided for in the AI Act.
One of the main changes affects AI systems classified as high risk. These include certain tools used to select candidates, make decisions on working conditions, assess performance or manage workers.
The specific obligations for these systems will apply from 2 December 2027. In the case of high-risk systems integrated into certain products, the date is postponed until 2 August 2028.
However, this postponement does not generally affect the transparency obligations that start to apply on 2 August 2026.
The Omnibus Regulation also modifies the AI literacy regime. Companies must adopt measures to support the training of people who use these systems, considering their experience, their functions and the context of use.
They are no longer required to guarantee a specific level of knowledge. However, they must promote appropriate and responsible use of AI.
What should companies review?
Although some obligations for high-risk systems have been postponed, companies should not delay the review of their artificial intelligence tools.
For example, before 2 August, companies should check which AI systems they currently use, which employment processes they are involved in and whether they interact with candidates, employees or third parties.
The first step is not only to update texts or add notices. It also requires understanding which tools are actually being used within the company and establishing clear criteria for their incorporation.
