EU AI Act transparency requirements: What small businesses need to know

The transparency requirements of the EU AI Act will apply from Sunday, 2 August 2026. In practice, they are intended to ensure that people know when they are interacting with AI or viewing certain types of AI-generated or AI-manipulated content.

For small business owners, the key questions are straightforward: does an AI chatbot on a website need to be labelled, and does every AI-generated image need an AI label?

The short answer is that a chatbot must disclose that it is AI. However, an ordinary AI-generated illustration does not need a visible label simply because it was created using AI. Deepfake content is a different matter.

AI chatbot users must know they are interacting with AI

Under Article 50 of the AI Act, providers of AI systems intended to interact directly with people must design their systems so that users are informed that they are interacting with AI. This information must be provided clearly, no later than at the time of the first interaction. The exception is when the use of AI is already obvious to the user.

In practice, website visitors should not have to guess whether they are receiving a response from a person or from AI.

Groweo’s AI applications account for this transparency requirement: visitors can tell from the beginning of the conversation that they are interacting with AI. As a Groweo customer, you do not need to build a separate chatbot disclosure.

Do AI-generated images need to be labelled?

Not automatically.

The AI Act distinguishes between the provider of an AI system and the business using it. Providers of AI systems that generate images, audio, video or text must ensure that their outputs are marked in a machine-readable format so they can be detected as artificially generated or manipulated.

This technical obligation rests with the system provider. If your small business uses an ordinary illustration created with an AI image generator on its website, in its marketing or in an application installed on its website, Article 50 does not generally require you to add a visible “AI-generated” label to the image.

European Commission guidelines confirm the same division of responsibilities: providers are responsible for machine-readable marking of AI-generated content, while disclosure obligations for businesses using AI systems apply to content such as deepfakes and certain texts published to inform the public on matters of public interest.

Deepfakes must be disclosed as AI-generated

If an AI-generated or AI-manipulated image, audio recording or video is a deepfake, the business publishing it must disclose that the content has been artificially generated or manipulated.

A deepfake is not limited to a fabricated video of a well-known person. The AI Act’s definition also covers content that resembles an existing person, object, place, organisation or event and would falsely appear to a person to be authentic or truthful.

A safe rule of thumb for small businesses is therefore this: you can use AI-generated illustrations without a separate visible AI label, but you should not publish misleading AI-generated content depicting real people, places, businesses or events. If you use deepfake content as defined by the AI Act, its AI origin must be clearly disclosed.

What about a blog post written with AI?

The AI Act also requires disclosure of AI-generated or AI-manipulated text if it is published to inform the public on matters of public interest. However, disclosure is not required if the content has undergone human review or editorial control and a person or business holds editorial responsibility for its publication.

Using AI as a writing aid does not remove a business’s responsibility for the accuracy of the text it publishes.

What changes for Groweo® customers?

You do not need to change how you use Groweo’s AI applications because of the chatbot transparency requirement that applies from 2 August 2026. The applications clearly inform visitors that they are interacting with AI.

Groweo applications, including Billboard content, can also continue to feature AI-generated images. Ordinary illustrations do not need a visible AI label simply because they appear in an application. The same applies to AI-generated images used elsewhere on your website or in your marketing.

You should, however, avoid deepfake content in all contexts. If an AI-generated or AI-manipulated image used in an application, on a website or in marketing material meets the definition of a deepfake, its AI origin must be clearly disclosed.

Please note that this article provides a general overview of the transparency requirements under Article 50 of the EU AI Act from the perspective of Groweo applications. If you are unsure about content used in your own marketing, consult the sources below or seek advice from a qualified expert 🧡

Sources