The AI Act and its impact on the use of chatbots

A central point of the AI Act is the classification of AI systems according to their risk potential. This takes into account the extent to which an AI system interferes with people’s daily lives and the potential dangers associated with it.

For chatbots, the “limited risk” category is particularly relevant. This means that chatbots are generally classified as limited risk applications. This has some important implications for chatbot operators.

From 2 February 2025, AI systems with unacceptable risk will be banned in the EU. These include technologies that manipulate or influence people subconsciously, as well as social scoring as used in authoritarian states. Predictive policing, which is based solely on profiling, is severely restricted as it is considered discriminatory and is scientifically controversial.

Real-time biometric surveillance in public spaces, in particular facial recognition, is also prohibited – with the exception of criminal prosecution for serious offences. While the bans mainly focus on government agencies and authorities, financial institutions must also ensure that their AI systems do not violate the new regulations.

The AI Act does not only affect Europe: several US states and Asian countries are adopting similar regulations in order to create standardised international legislation. This should facilitate the global use of AI and minimise risks at the same time.

What does this mean for your company?

There are no comprehensive restrictions for companies in the first phase. However, stricter rules for high-risk AI will apply from August 2026, including logging and reporting obligations. A general ban on such applications is not planned.

One of the central requirements of the AI Act is the creation of transparency in dealing with AI systems. In the case of chatbots, this means that it must be clear to users that they are talking to a bot and not a human. This is to ensure that users understand the nature of the interaction and can respond appropriately.

Another important aspect is that users must be able to cancel interactions with a chatbot. This is an important safeguard to ensure that users remain in control of the interaction.

It is important to note that the regulation of the AI Act affects all market participants who use chatbots. This means that companies and organisations will be uniformly required to design their chatbots in a transparent and ethical manner. This serves to ensure that chatbots are used in a way that is fair to users and does not pose unnecessary risks.