AI Act in force: Is your website compliant?

Yesterday, August 2, 2026, the transparency obligations of the AI Act came into force. Until now we have been preparing to comply, but now we need to check whether we actually do.

In this article, we won't review again what the AI Act is or all of its obligations. Instead, we will focus on something urgent for any online store: how to detect whether your e-commerce website is among those that comply or among those exposed to a penalty.

1. The deadline has arrived

While the AI Act was pending entry into force, many companies treated compliance as a pending project, something to review. That window has now closed. From now on, any competent authority can open a case, and any customer can file a complaint for not having been informed that an image, a video, or a conversation were generated by AI.

Right now, failing to comply with AI labeling on your website is a violation that is already occurring every minute the image remains published without the corresponding notice.

2. The quick self-assessment: is your website non-compliant right now?

Instead of going through the entire list of obligations under the AI Act, ask yourself these four concrete questions:

• Can a visitor identify, without having to search, which images in your catalog are generated or substantially altered by AI?
  Remember that it's not enough for the information to exist in your legal notice—it must be visible at the moment the content is displayed.

• Do you know for certain whether your stock images or images from external suppliers contain AI-generated content?

• Does your chatbot or virtual assistant make it clear, in its first message, that the user is talking to an AI?
  During the first interaction with the chat, it must be apparent that the user is interacting with a machine and not with a person.
  
• Can you demonstrate the measures you've taken to identify AI-generated content appearing on your website?
  It's not enough to comply; you bear the burden of proving that you've taken the necessary measures to comply if an authority requests it.

If you've answered "no" or "I don't know" to any of these questions, it's highly likely that your website is failing to meet the transparency obligations of article 50 of the AI Act.

3. What to do if you detect that you're non-compliant

If after the self-assessment above you detect that your website has shortcomings, the most important thing is that you take steps to comply:

1. Correct what is most visible to users first. The first message of the chatbot and product images that constitute a deepfake are the points where a complaint is most likely.
2. Document the date on which you apply each correction. The AI Act views cooperation and proactive correction of violations positively when determining a possible penalty.
3. Review the origin of your third-party content. It's likely that your website incorporates images that your suppliers have provided, and you may not know if they have been generated or altered by AI. Therefore, it is very important that you review and verify their origin.
4. Continuous verification process. The AI Act does not require labeling to be done once, but rather a continuous process, because website content changes constantly.

4. How does it affect the legal compliance of my e-commerce website?

For an e-commerce, complying with the AI Act is a requirement that involves your entire website because you must review your image catalog, your customer service chatbot, your product search engine if it uses an AI-trained algorithm, whether you use virtual try-ons so customers can test products, etc.

Unlike other regulations, in this case any customer can verify for themselves whether a website labels its images or whether its chatbot identifies itself as AI. This not only can pose legal risks, but also reputational risks because customers can easily verify that you are not complying with European AI regulations.

Therefore, for an e-commerce, this compliance must be integrated as part of continuous website maintenance.

Don't leave your e-commerce exposed and comply today

With the AI Act already in force, your e-commerce's legal exposure depends on decisions you can make right now. At Lawwwing we help you solve it with two solutions that work together.

AI Sentinel periodically scans the visual content published on your e-commerce website and analyzes it with its own detection engine, trained from millions of images. Its technology allows it to differentiate a real image from one generated with AI, with reliability exceeding 98.5%.

For us, complying with the law and caring for your business are not incompatible things. Our goal is to ensure that disclosing AI use does not harm your sales, so AI Sentinel does not overlay labels on your product photos. Instead, it displays that information through a floating, accessible widget that respects your store's design and does not interfere with the shopping experience.

Additionally, if you need to verify images before uploading them, check material sent to you by users or suppliers, Lawwwing puts Vericta at your disposal—an image and video verification API that uses that same detection technology with identical precision.

With the AI Act already in force, don't wait any longer and hire Lawwwing.