Two laws took effect this year that most companies making video content haven't fully registered yet.
On June 9, 2026, New York became the first state in the country to require conspicuous disclosure whenever a commercial advertisement features an AI-generated "synthetic performer." On August 2, 2026, the EU AI Act's Article 50 became applicable, requiring visible labeling of AI-generated video, audio, and images, with an even stronger disclosure duty for anything that could pass as a real person. Both apply regardless of where your company is headquartered — if your content reaches a New York or EU audience, you're in scope.
If that's the first you're hearing of either law, you're not alone. And if your first reaction is a flicker of worry — wait, are we already out of compliance? — that's the objection this post is here to answer.
It's a fair question, and the honest answer is: it depends entirely on how you do it.
The data on AI disclosure and trust is genuinely mixed. Some research shows people trust a disclosed ad more than an undisclosed one. Other studies show that simply slapping an "AI-generated" label on content makes people rate it as less natural, even when nothing else about the content changed. Consumers overwhelmingly say they expect disclosure — but expecting it and being reassured by it are two different things.
Here's what the mixed data actually points to: disclosure isn't a trust hack. It's not a badge that makes people like your content more. What builds trust is what disclosure is paired with — whether the content still feels considered, crafted, and accountable to a real person who stands behind it. The brands that get burned aren't the ones that disclose. They're the ones that get caught not disclosing, after the fact, looking like they had something to hide.
That's a different problem than the one most companies are trying to solve. It's not "how do we avoid disclosure." It's "how do we make disclosure something our content can absorb without looking diminished by it."
We label our AI-generated work by default — visibly, consistently, on every piece, whether a client asks or not. It's not a response to New York or the EU. It's been our production standard since before either law existed, because the same conviction that shapes everything else we make — that directing craft and narrative judgment can't be commoditized — also means we never wanted to hide which parts of a piece a human directed and which parts AI executed.
That's the whole idea behind Humans Direct. AI Executes. A director is still in the room. An editor still signs off. The AI handles the parts that scale — the languages, the variations, the volume — but a person is accountable for the story, and the disclosure reflects that honestly instead of apologizing for it.
Practically, this means:
For a company that's spent months building trust with an audience, that difference is the whole game. The two laws didn't change how we work. They just gave a name to why it was already the right way to work.
Training and onboarding video. A digital twin of a subject-matter expert delivers a compliance module in twelve languages. The disclosure sits quietly in the corner of every version — consistent, expected, never a surprise to the learner halfway through.
Customer-facing marketing. A product explainer built around an AI avatar ships with the same visible disclosure a human-shot video would carry as a production credit — treated as part of the craft, not a caveat squeezed in by legal.
Executive and internal communications. A CEO's digital twin delivers a quarterly update to a distributed workforce. Employees know exactly what they're watching and why it was made that way — which, for an audience already primed to be skeptical of anything that feels manufactured, is precisely the point.
The laws are new. The standard doesn't have to be.
