Publishing AI-assisted content without a disclosure could put your business out of compliance with new EU regulations—even if your business is based in the US. The EU Artificial Intelligence Act is rolling out in phases; it started in August 2024 and will take full effect in August 2026. If your content reaches readers in the EU, these rules may apply to you, and communications teams and authors need to act before the deadline.
What Is the EU Artificial Intelligence Act?
The EU AI Act is a large-scale attempt to regulate AI. The act sets expectations around transparency and accountability regarding AI-generated content—particularly for content meant for public consumption. The act applies based on where content is consumed, not where it’s created. That means the regulations may apply to writers and content teams outside the EU if their content reaches audiences in the EU.
The regulations are aimed at two groups: providers and deployers. A provider is a company that builds AI systems—think Google or Perplexity. Most of the direct disclosure obligations fall on providers of certain AI system types.
Businesses, agencies, or professionals who use AI tools and publicly share their output are deployers. Deployers have obligations primarily in high-risk contexts, but the regulatory guidance around deployer obligations is still evolving. This article focuses on the act as it currently applies to deployers.
As a small-business owner who uses online marketing and AI tools, I count as a deployer. My content is available to readers in the EU, and that means the EU AI Act applies to me, even though I’m based in the US.
Writers, take note here: Let’s say you don’t use AI tools, but your assistant, researcher, editor, publisher, or marketing team uses AI as part of their workflow. If your published content is available in the EU, then the act still applies to you too.
Let’s break down what the act requires from deployers.
Key Points of the EU Artificial Intelligence Act
The EU Artificial Intelligence Act has three key points writers should be aware of:
- If you use AI tools, publicly distribute their output, and the content reaches individuals in the EU, you may be responsible for disclosing your use of AI.
- The act doesn’t ban the use of AI tools. This is about making sure professional and commercial users are transparent and accountable.
- AI-generated, public-facing, informational text presented as factual, authoritative content may require disclosures or labels—when it’s published without meaningful human oversight.
Content teams, that third point is crucial. Any news-style content, health guidance, financial information, or policy commentary that is generated by AI without meaningful human oversight needs a label or disclosure.
The act doesn’t ban the use of AI tools, and people don’t need to disclose every AI experiment or casual use. You can still use AI to help with drafting, if you choose.
The EU Artificial Intelligence Act regulations apply to professional, commercial use of AI-generated content that’s publicized without clear editorial control or human responsibility.
Understanding what triggers the act’s requirements is only half the picture—the other half is knowing how your editorial process can fulfill them.
How the EU Artificial Intelligence Act Affects Writers
Human editorial control affects content, and the act recognizes that. If your AI-generated content goes through a substantial human review, that shifts the responsibility to the human or organization acting as the author. The act treats that content as human-led, not machine-produced. So what counts as a substantial human review?
- Fact-checking: Confirming sources, statistics, assertions, etc.
- Rewriting: Revising text for clarity, tone, or flow
- Judgment calls: Evaluating whether a claim needs a source or whether a section should be cut
- Final approval: Deciding whether, when, and how a piece should be published
What doesn’t count as a substantial human review?
- Light proofreading
- Simple grammar revisions
The difference lies in real editorial responsibility. We can use AI to help us create content, but we are responsible for that published content. A human must read the text, evaluate it, and approve it for publication. Regulators are still working out the details on this in a Code of Practice, and it should be finalized in June 2026.
EU Artificial Intelligence Act Checklist for Content Teams
Nonfiction authors and B2B communications teams—take note of the changes the EU Artificial Intelligence Act requires and update your publishing processes. Use this checklist to make the transition easier:
Evaluate Current Processes
Start by monitoring how you use AI in your writing process now and think about how you plan to use it in the future. Note which tools you use that incorporate AI already. You’ll need that information for your AI policy.
Formalize Your Workflow
Make it clear how your writing and editing processes work. Document the standards you use for drafting, editing, and publishing, and state who is responsible for what. This includes training to ensure that your staff has a sufficient level of AI literacy. Clear documentation can help streamline your processes as your team members change and act as a legal safeguard.
Draft an AI Policy
Having an official AI policy for your business demonstrates transparency and responsible business practices. It tells your readers, clients, and partners in what situations you will or won’t use AI and outlines the tools you may use.
Pro tip: Once the policy is in place, use it to help make decisions about updates to your content-creation process. Is a new tool or action you’re considering in line with your policy?
Create a Set of AI Disclosure Templates
Disclosures proactively communicate relevant information, providing readers assurance that they can rely on the content. It’s safe to trust that the content is authentic. Without disclosure, readers don’t have a basis to make an informed decision whether to trust the content. Depending on how you work, you may need disclosures for multiple points of the writing process
- When an author submits work to their editor, the author should disclose if and how AI assisted them with their work. “I used Claude to organize my outline.”
- Before an author and editor sign a contract, both parties should disclose how they may use AI in their roles. “I use Perplexity to create a list of keywords.”
- When an editor returns a revised file to an author, the editor should disclose if and how AI tools assisted them with their work. “I used ProWritingAid to identify echo words before the first pass through the manuscript.”
- When a piece is published, the author or publisher should include a disclosure to readers explaining how AI was used. “This article about dog training was written and edited by humans with the assistance of AI tools for minor outline, revision, and formatting purposes.”
Save time by creating a template for your disclosures. You can update the specifics on a per-project basis, but you’ll have the boilerplate language ready when needed.
Need an in-depth discussion of AI disclosures or a sample policy? I've got you covered. In May 2025, I wrote about AI and ethics, developed an AI policy, and created my AI disclosures notice.
Let’s Team Up
August 2026 isn’t far off. If you’re a nonfiction author or content team working with AI tools, now is the time to get your policies, workflows, and disclosures in order—not after the deadline. The good news: You don’t have to navigate this alone. I track regulatory changes, editorial standards, and industry shifts so you can stay focused on your work. Subscribe to my monthly newsletter and get plain-English updates delivered to your inbox every month—no legalese, no overwhelm.
When you’re ready for expert eyes on your content, I’m here to help protect your brand and set your writing apart. Together, we can make your writing bulletproof!
Full disclosure: This article was written by a human and revised with the assistance of AI tools. The final editorial decisions were made by a human editor to ensure the content meets my quality standards.