Labelling AI-made ads: one US state wrote a law, everyone else can build a habit

Pennsylvania now requires a label on advertising that AI has significantly modified or generated, in the same format as the ad itself, for large companies and small businesses alike (The Verge). The text already raises questions: who decides what is "significantly modified", and what counts as "evidently fictional", which is exempt?

Other countries took the opposite route. In May 2025 the Swiss National Council rejected a specific deepfake law by 111 to 70, arguing that criminal and civil law already cover the gaps (parlament.ch). What exists there instead is voluntary: a "Created with AI" label for ads on the country's largest marketplaces, free for any company to use (netzwoche).

Side by side, the lesson is universal. A law has to define the edge cases, and that is hard. A habit does not. A company that labels its AI-made visuals, voices and texts clearly and early never needs to argue about definitions, and it earns the trust that laws try to force.

Behind every AI-made ad is also a designer, copywriter or voice actor whose work changed overnight. Many of them do not need a new career, they need a few weeks of training to direct the tools instead of competing with them. That is the kind of training your donation pays for.

Practical step: decide which AI-made content you label, use one recognisable label everywhere, and write the rule down.