Should You Disclose AI Use? The Trust Question

This stopped being a purely ethical judgment call in 2026. On August 2, 2026, two of the world's most consequential AI transparency regimes became simultaneously enforceable: California's SB 942 (the California AI Transparency Act) and Article 50 of the European Union's AI Act. For the first time, AI disclosure moved from "something to consider if a reader complains" to a legal obligation with real financial teeth in two major jurisdictions at once (2026 analysis of the simultaneous California and EU AI disclosure enforcement). That doesn't mean every blogger is suddenly a regulatory target — but it does mean the ground under this question has genuinely shifted, and it's worth understanding exactly what changed before deciding what to do about your own blog.

What the new laws actually require — and who they're actually aimed at

It's important to be precise here, because a lot of coverage blurs the distinction between rules aimed at AI companies and rules aimed at individual publishers. California's SB 942 applies specifically to large generative AI providers with over one million monthly users — meaning the companies building tools like ChatGPT or image generators, who must now offer free AI-detection tools and embed both visible and machine-readable disclosures in AI-generated output. It does not, on its face, require an individual blogger to label every AI-assisted sentence (the same 2026 disclosure law analysis).

The EU's Article 50 is broader in scope but similarly aimed primarily at "providers" of generative AI systems, who must mark synthetic outputs in machine-readable form. Where it does reach individual publishers directly is as a "deployer": if you publish a deepfake, or certain AI-generated text specifically on a matter of public interest, without disclosure, you may have direct disclosure obligations of your own — though the exact scope depends on your role, the content type, and how much human editorial oversight was involved (2026 legal analysis of EU AI Act Article 50 obligations). At the U.S. federal level, a separate AI Labeling Act and a bipartisan deepfake-labeling bill have both been introduced in Congress as of 2026, but neither has been enacted — meaning there is currently no standalone federal AI disclosure statute in the U.S., regardless of state-level rules like California's (the text of the proposed federal AI Labeling Act of 2026).

The rule that already applies to you, right now, regardless of any new law

Here's the part most disclosure discourse skips, and it matters more for the average blogger than any of the new AI-specific legislation: the FTC's existing Endorsement Guides and Section 5 of the FTC Act already apply, and always have, to any blogger publishing reviews, recommendations, or firsthand-experience claims — with no need for new AI-specific legislation to trigger them. Under these existing rules, an endorsement must reflect the genuine opinion of the endorser, material financial connections must be clearly disclosed, and — critically — you cannot describe a product experience you didn't actually have (Dynamis LLP's 2026 legal analysis of AI disclosure obligations for brands and creators).

That last point is the one worth sitting with. If AI drafted a passage that reads as "I tested this product for six weeks," and you didn't actually do that testing, you're not just risking reader trust — you're in the territory existing FTC deception rules were built to address, regardless of whether any AI-specific law applies to your situation at all. This existing framework, not the newer AI-specific statutes, is the most directly relevant legal reality for most individual bloggers today.

What bloggers are actually doing, and the gap that creates real exposure

Set against this legal backdrop, current blogger behavior looks riskier than it did before these laws existed. A 2026 survey of 81 active bloggers found nearly half say the majority of what they publish is AI-assisted, including 31.3% who said 76–100% of their content is. Of those heaviest AI users, 84% never disclose it to readers, and across bloggers overall, only 10% say they always disclose AI involvement (RightBlogger's 2026 State of Blogging survey). None of that behavior was illegal before August 2026 in most cases, and much of it likely still isn't under the new laws' narrower scope — but the FTC's existing endorsement rules mean the riskiest subset of that behavior (claiming firsthand experience you didn't have) was never actually safe, new legislation or not.

A practical disclosure framework that accounts for both law and trust

Given all of this, a workable policy separates three distinct questions rather than treating "disclose or don't" as one binary choice:

• Is this content plausibly covered by an actual disclosure law? If you publish AI-generated or AI-manipulated images, audio, or video, or you have any EU or California audience exposure and publish AI-generated text on public-interest topics, it's worth understanding your specific obligations under Article 50 or SB 942 rather than assuming they don't apply — a short consultation with a lawyer familiar with the current rules is cheap insurance if this describes a meaningful share of your content.
• Does this content make a firsthand-experience or endorsement claim? If a post says or implies "I tried this," "I tested this," or "in my experience," and AI substantially originated that claim without you actually having the underlying experience, treat this as the existing FTC endorsement framework's territory — disclosure isn't optional risk management here, it's the difference between an endorsement and a fabricated one.
• Is this content mechanical assistance rather than either of the above? Using AI to tighten grammar, brainstorm headlines, or restructure a draft you actually wrote sits in a different, much lower-risk category under both the new laws and existing FTC guidance, and a blanket, once-stated AI-use policy (rather than a disclaimer on every post) is a reasonable way to handle it.

Why disclosure might be a competitive advantage, not just a compliance cost

There's a version of this that isn't purely defensive. As the regulatory environment tightens and reader skepticism of undisclosed AI use grows, a blog that's already transparent about its process has a genuine head start — both legally, since transparent practices are easier to defend if a law's scope turns out to apply more broadly than expected, and reputationally, since a clearly stated, honest AI-use policy signals exactly the kind of trustworthy, accountable authorship this series has argued readers are increasingly filtering for.

Frequently asked questions

Do the new California and EU laws actually apply to a typical individual blogger?
In most cases, not directly — both laws are primarily aimed at large AI system providers, with individual publisher obligations under the EU rule limited mainly to deepfakes and certain AI-generated public-interest content. But if you publish AI-generated images, audio, or video, or have meaningful EU or California readership, it's worth confirming your specific exposure rather than assuming the laws don't apply.

Is it actually illegal to publish AI-assisted content without disclosing it?
Generally, no, for ordinary AI-assisted text content outside the specific categories the new laws cover. The clearer legal risk comes from existing FTC endorsement and deception rules, which prohibit describing a firsthand experience you didn't have — a rule that predates and operates independently of any AI-specific legislation.

What's the single highest-priority thing to fix if I've never thought about this before?
Audit your posts that make firsthand claims — "I tested," "I tried," "in my experience" — and confirm each one reflects something you actually did. That's the specific area where existing law, not just reader trust, creates real exposure.

Should I add a blanket AI-disclosure statement to my blog regardless of legal requirements?
It's a reasonable, low-cost step given both the tightening regulatory environment and the reader trust research covered elsewhere in this series — a short, honestly-stated policy on your About page costs little and heads off a question a growing share of readers are actively wondering about.