← All writing

Two AI content disclosure laws took effect in 2026: Turkey's rule and the EU AI Act's Article 50, which requires conspicuous disclosure of AI-generated or synthetic content. Neither law treats a disclosure buried in a bio, a terms page, or a platform's built-in AI label as sufficient. The disclosure has to travel with the post itself, because the post is what people actually see, out of context, in a feed, a screenshot, or a repost.

That single requirement, disclosure that survives being seen alone, is the one most teams publishing across more than one account are quietly failing right now.

What do Turkey's rule and Article 50 actually require?

Both rules landed within a day of each other in 2026, on two different continents, and both point at the same underlying problem: it's no longer enough that AI involvement is disclosed somewhere. It has to be disclosed with the content, in a form a viewer encounters without having to go looking for it.

I went through my own disclosure practice against that bar, expecting a quick check. It was smaller than I expected, and more irritating than I expected. The easy question, whether anything I publish is AI-assisted, answers itself. The narrower question is the one that matters: if someone sees a single post, out of context, with no profile visit and no scroll, do they know?

Why doesn't a bio-line disclosure count?

Because it doesn't travel. A line in a bio discloses to people who read bios. A platform's built-in AI label discloses to people who open the right menu. A terms page discloses to people who click through to it, which is close to nobody. None of those three things are attached to the post when the post leaves the platform it was written for, which is most of the time: it gets screenshotted, quoted, reposted, pulled into a roundup, or simply seen by someone scrolling past who never visits a profile.

The regulatory language ("conspicuous," "travels with the content") is really describing a design constraint, not a policy requirement. If the disclosure only exists one click away from the post, it functionally doesn't exist for most of the audience that sees the post.

How do you fix it across multiple accounts?

The fix isn't clever, and it isn't a tool. It's a decision, made once per surface: what does disclosure look like on this platform, on this post format, when the post is fully separated from everything around it? Then apply that decision the same way every single time.

Written down, that's a paragraph. Doing it consistently across a set of accounts is a different job, because publishing tools are very good at multiplying whatever practice you already have, sloppy version included. A team running five client accounts with an inconsistent disclosure habit doesn't have one small gap. It has five gaps, compounding every time a post goes out.

This is the part of the AI-disclosure conversation that gets skipped in most commentary about the new rules: the rules are actually easy to understand. The failure isn't comprehension. It's operational consistency at the moment a post is scheduled and sent, across every account that publishes AI-assisted content, every time, without relying on someone remembering to add the line manually.

The question worth asking your own team

If you publish anything AI-assisted across more than one account: where does the disclosure actually live, and does it travel with the post, or only with the profile?

That's the audit worth running before a regulator, a platform, or a client runs it for you.