— Perspectives
AI watermarking is coming for every image you publish. Get ahead of it.
Google's SynthID provenance system is being adopted across the industry to invisibly flag AI-generated content. If your brand uses AI imagery — and most now do — labelling is about to become visible to your audience.
There is a quiet shift happening in how the internet will treat AI-generated content, and it will reach brand marketing faster than people expect. Google's SynthID — a system that embeds invisible provenance signals into AI-generated images, audio and other media — is now being adopted by a widening circle of major AI companies, and detection is expanding into mainstream products people use every day.
The plain-language version: the tools your audience already uses are getting better at quietly knowing what was made by AI, even after edits. Provenance is becoming infrastructure, not a niche concern.
For brands, this lands on a slightly awkward truth. A great deal of marketing imagery is now AI-assisted or AI-generated. That is not a scandal — it is a sensible use of the tools. But "nobody can tell" is quietly ceasing to be true, and the brands that will look worst are the ones who leaned on AI imagery while implying everything was shot, hand-made, or real.
The move here is not to stop using AI. It is to decide your posture before the platforms decide it for you. Where does AI imagery genuinely serve the work, and where does authenticity actually matter to your audience — real product shots, real people, real places? A brand that is deliberate about that line has nothing to fear from provenance labels. A brand that blurred it to cut corners is about to have that visible.
Transparency is going to stop being a values statement and start being a technical fact attached to the file. Better to own your posture now than to have it revealed later.
← All signals