EU AI Act news and IAB’s AI Accountability for Publishers Act

  • 20 / 02 / 2026
  • Alicja Graczyk
EU AI Act news and IAB’s AI Accountability for Publishers Act

As AI moves from a “wild west” tool to a regulated industry standard, digital publishers are finding themselves at a crossroads. With the full implementation of the EU AI Act on the horizon, the distinction between “using” AI and “owning” it has never been more critical. From the truth about who actually owns those AI-generated images to why labeling AI content is starting to become essential, we’re breaking down what’s changing and why your human touch is still your biggest legal superpower!

source: https://giphy.com/

What is EU AI Act

Officially, the EU AI Act is “a European Union regulation concerning artificial intelligence (AI)”. Its top priority is ensuring AI is reliable and doesn’t discriminate. European Parliament and Council want to be able to trace how an AI makes decisions and ensure it’s not hurting the planet. Their golden rule? A real person should always be supervising the tech to prevent harm.

When it comes to the timeline, the new AI rules officially started in August 2024, but the EU is rolling them out in stages over three years to give everyone time to get ready. For instance, the most dangerous types of AI were banned back in February 2025, and by August 2025, companies making tools like ChatGPT had to become more transparent about how their tech works. For most people running websites or online businesses, the biggest date to remember is August 2026. That is when almost all the rules will finally apply to everyone.

EU AI Act and digital publishers

According to Article 50 of the AI Act, if an artificial intelligence helped you make a deepfake (highly realistic but fake video, audio, or photo material) that could pass for reality, you’re legally required to flag it. If you’re publishing something clearly artistic or a piece of fiction, it’s still best to mention the use of AI, but you can do it in a way that doesn’t distract from the work itself (for instance, not in the form of a clear watermark, but rather mention in the description).

To sum up: while most headlines focus on tech giants like OpenAI or Midjourney, by August 2, 2026, smaller publishers and site owners will officially step into the role of “deployers.” This brings some specific homework to your desk:

  • Labeling is a must: any AI-generated text, image, or video that “informs the public on matters of public interest” has to be clearly marked. The only loophole? If the content has undergone proper human editing, with a person taking full editorial responsibility for the final result;
  • Techy standards: you’ll need to make sure these AI labels are “machine-readable.” This means the info shouldn’t just be in the caption; it needs to be baked into the file’s metadata so search engines and filters can spot it instantly;
  • Handling deepfakes: if you’re posting AI-tweaked content that looks like a real person or event, you’ve got to be upfront about it the second a user sees it, think “AI” icons or a very clear disclaimer right at the start.

IAB draft legislation

The Interactive Advertising Bureau (IAB) is taking a stand against what they describe as an “existential crisis” for digital media by proposing the AI Accountability for Publishers Act. Announced at their 2026 Annual Leadership Meeting, the draft legislation targets the large-scale scraping of publisher content by AI companies that use this data to train models and generate summaries without compensation.

Commercial use of content from ChatGPT, Gemini, and Midjourney

In 2026, digital publishers can definitely use content from ChatGPT, Gemini, and Midjourney for commercial projects, but “using it” and “owning it” are two very different things. While most paid subscriptions give you the green light to monetize what you create, there’s a massive catch: in most cases, you don’t actually own the copyright. Current legal standards in the U.S. and the EU are pretty clear that AI-only work can’t be copyrighted. So, while you can hit “publish” and make money, you might find it impossible to stop a competitor from lifting your AI-generated articles or images and using them for themselves.

From compliance to monetization

Ultimately, these new rules aren’t just about avoiding an EU fine; they are about protecting your ad revenue. As we move deeper into 2026, the industry is shifting toward brand safety as a top priority. Why it’s important for monetization: advertisers are increasingly wary of where their budgets go, and many can now block ads from appearing on sites that infringe on copyright or fail to clearly mark obvious AI-generated content. By being transparent and including a human in the process, you aren’t just staying legal, you’re ensuring your site remains a “safe bet” for the premium advertisers!

* Please consider this a helpful guide, not a legal consultation. We’re sharing this for educational purposes only. Since laws change faster than your feed, you should definitely speak with a qualified attorney before making any big moves.

Read also

Google Magi
Google Magi

Google Magi

Curious about the future of publishers, particularly in light of Google's AI initiatives? Look no further! Discover more about Google's secretive project, Magi!

Read more
FT 1000: optAd360 is one of the fastest-growing companies in Europe
FT 1000: optAd360 is one of the fastest-growing companies in Europe

FT 1000: optAd360 is one of the fastest-growing companies in Europe

We are listed among a thousand European companies that experienced rapid growth in the past years. Come and celebrate with us!

Read more

Find the best solutions
for your business

Benefit from expert knowledge

Start earning more

Registration to the optAd360 network

Increase your ad revenue!

Join satisfied publishers who, thanks to the optimization of their advertising space
with our technology, started to generate greater profits.

Sign up