Article 50 of the EU AI Act became enforceable on August 2, 2026. If your site runs a chatbot or a live-chat widget with AI replies, the transparency duties already apply to you. Furthermore, this applies to AI-generated or AI-edited images that could pass as real; these duties apply from three weeks ago, not a future date.

That’s not a scare tactic. In short, it’s the calendar.

National market surveillance authorities can enforce Article 50 starting now. In fact, these fines can reach €15M or 3% of global turnover. As such, this applies to companies large enough for that ceiling to matter.

According to the European Commission’s transparency rules fact page, these obligations became enforceable on August 2, 2026, with penalties reaching up to €15 million or 3% of global annual turnover.

In addition, for everyone else, the more immediate risk is simpler. Specifically, this risk involves an authority, a competitor, or a customer complaint. Ultimately, such an event surfaces a site with no disclosure at all.

This kind of enforcement pressure is exactly the backdrop that business strategy for uncertain economic times is meant to address. Indeed, regulatory risk can now hit as fast as a competitor’s complaint.

What Does Article 50 Actually Require?

Two duties cover most WordPress sites:

According to the European Commission’s official FAQ on Article 50, these transparency duties cover disclosing AI interactions, marking AI-generated content, and labeling deepfakes and AI-generated public-interest text.

Chatbot disclosure (Art. 50(1)). If visitors can reasonably mistake your chatbot, voice agent, or AI-powered live chat for a human, they must be told this. Specifically, they need to know they are talking to AI; this disclosure must happen clearly at the start of the interaction. Notably, a line buried in your terms of service does not satisfy this. In fact, the European Commission’s July 2026 guidance is explicit. This guidance states the disclosure needs to sit where the interaction happens.

AI image labeling (Art. 50(4)). Images that are AI-generated or AI-manipulated need labeling. They must be labeled as such if they could appear authentic. This label must be visible to a visitor or readable by a machine.

It does not require labeling for most site owners. This applies to ordinary AI-assisted marketing copy or product descriptions. However, a human editor must take responsibility for the content.

The text-labeling duty is narrower than people assume. It specifically targets AI-generated text. This text must be presented as information on matters of public interest.

Sorting out which parts of a workflow actually need disclosure is itself a small exercise in how businesses navigate market disruption. Notably, the rules apply unevenly across genuinely similar-looking tools.