New AI transparency requirements have been enacted, and they land squarely on two groups: advertisers and deployers — the companies that put AI systems into real-world use rather than build them.

That's the headline. Behind it sits a familiar pattern. Regulators keep circling the same question: when people interact with, or see the output of, an artificial intelligence system, do they know it? These new requirements say they should — and they put the legal burden on the firms that monetize or operate the technology.

Who has to do what now?

The rules target advertisers and deployers of AI systems. For advertisers, the practical effect centers on disclosure: audiences must be told when content they're seeing was generated or shaped by AI. For deployers — companies running AI tools in live services rather than developing them in-house — the obligations extend to how those systems are presented to end users.

Transparency, in this context, means the user-facing surface: labels, notices, and disclosures that make clear an AI is involved. It's the same principle that already governs paid promotions and sponsored content. Now it extends to synthetic output.

The takeaway: if you run ads or deploy AI-facing products, disclosure is no longer optional.

Why does this matter commercially?

Compliance work costs money. Teams that buy advertising or operate AI-driven features now need processes to flag AI-generated material, apply the required notices, and document that they did so.

For agencies and ad-tech intermediaries, the rules create a chain-of-custody problem. A brand that runs an AI-generated campaign needs assurance its partners applied the right disclosures. Expect contract language to shift, with disclosure obligations pushed down to vendors and platforms.

For deployers, the exposure is broader. A company that embeds a third-party AI system into a customer-facing product can't simply point at the vendor. The entity putting the system in front of the public carries the duty.

Plainly: accountability follows the deployment, not the codebase.

What's the wider signal?

Regulation of AI is moving from framework documents to enforceable duties. Early rulemaking focused on high-risk uses — hiring, credit, biometrics. Transparency requirements for advertising and deployment reach much further, touching ordinary commercial speech and everyday product interfaces.

That reach is the point. Disclosure rules are cheap to state, hard to dispute, and easy to audit compared with, say, assessing algorithmic bias. Regulators like them for exactly that reason.

Advertisers and deployers should read this as a floor, not a ceiling. Transparency duties tend to expand once the labeling infrastructure exists.

In short: the era of unlabeled AI output in ads and consumer products is closing, and the compliance bill now lands on operators.