OpenAI introduces text labeling to comply with the EU's AI Act

OpenAI is introducing text labeling meant to help make it clear when text has been made by AI. That's according to MLex, and the background is the requirements of the EU's AI Act.

For you as a private user, it hardly changes anything from one day to the next. However, you may eventually come across text labeled as AI-made, which can make it easier to see what you are reading.

For companies, the point is more important. If you use AI to write newsletters, website copy or customer emails, you should keep an eye on how the labeling works in practice and whether it follows the text when you copy it elsewhere.

When should you act? There is no reason to panic. Wait until OpenAI has described how the labeling works, and then check your own workflows. If you use AI for texts published under your name, it is a good idea to get the rules sorted out in good time.

What it means for you

If you use ChatGPT for texts, you may eventually find that they are labeled as AI-made, so you and others can more easily see what has been written by a machine. If you work with newsletters, website copy or customer emails, check whether the labeling comes along when you copy the text elsewhere. There is no reason to hurry, but when OpenAI explains how it works, you should review the texts published under your name or your company's name.