Let’s talk about the subscription you’ll soon need just to be taken seriously.
Right now, if you publish something true, the internet still assumes you’re probably real. That assumption is being switched off. Not in one loud announcement — the way most important things actually change, which is quietly, at the infrastructure, in defaults nobody reads.
Picture a photographer in Boksburg. She shoots a wedding, posts ten frames, and one of them gets flagged as possibly synthetic — because it carries no provenance metadata. She didn’t forge anything. She used a camera body that doesn’t sign its images and a laptop that strips the data on export. Nothing about her work is dishonest. It’s just unprovable. And in a system that’s tipping toward “unproven means suspect,” unprovable and dishonest start to look the same.
That’s the shift. And it isn’t coming from the fringe. It’s coming from the centre — from the companies that already own the pipes.
Cloudflare flipped a default so that machines it can’t identify get treated as trespassers, unless a site opts out. The payment networks spent the year building the rails that let an agent spend your money for you: a verifiable-intent layer here, a trusted-agent protocol there, a wallet standard somewhere else. Six of them, last count, and not one talks to another. The European Union is standing up a trusted-issuer list — the short list of companies allowed to speak for you, digitally, in law. And the content-credential standard that decides which photos and video carry a proof of origin is written by the same handful of platforms that own the feeds.
Read that as progress and it looks like the sensible answer to deepfakes. Necessary, even. Read it properly and it’s a pattern we’ve already watched four times — and the pattern doesn’t care about the reason. It just follows the money. Open by default, captured at the checkpoint.
Think about how the web got its toll roads the first time. Search was open until one company became the front door and started charging rent on it. Software was open until the app store became a gate and took thirty percent of everything that walked through. Email stayed technically open, and the ability to reach a stranger’s inbox became a business — sold back to you by the people who’d already hollowed it out.
Now the thing being tolled isn’t reach. It’s credibility. The right to be believed.
Here’s the mechanism, because this is the part people skip. You don’t capture trust by lying to people. You capture it by controlling the checkpoint where trust gets issued. Whoever decides which content counts as verified never has to make a single truth claim. They just move the default. Everything unsigned drifts toward the junk drawer — downranked, shadow-flagged, filtered, quietly invisible — and the pool of “verified” content becomes a moat they hold the keys to.
We already ran this experiment. Email ran it. After SPF, DKIM and DMARC went mainstream, unsigned mail didn’t get argued with. It just started landing in spam. Nobody ever stood up and said your message was false. The system simply stopped letting it through.
That’s the future being built right now for everything else. Photos. Video. Audio. Documents. Eventually your own voice on an outbound call. Not a verdict on whether you’re lying — a verdict on whether you bothered to prove you weren’t.
And look at who ends up holding the pen. The individual — the photographer, the podcaster, the founder recording a client call — has no signing infrastructure, no issuer relationship, no registry entry. The platform has all three, and sells them back to you as a feature. So the person with the least power to prove themselves is exactly the person who gets treated as the suspect.
That isn’t a fair fight. It’s a table where one side writes the standard and then sells the seats.
So what do you actually do about it, as an operator?
Stop treating verification as somebody else’s job. Own your keys. Sign your own work — timestamp it, hash it, keep the receipt somewhere you control, and put the proof where a sceptic can check it without asking your permission. If you publish a podcast, a video, a proposal, get in the habit of making a claim that travels with the file instead of one that dies the moment the file leaves the room.
And the less comfortable lesson: don’t rent your credibility. The blue check. The platform badge. The trusted-issuer status. Every one of them is a credential sitting on someone else’s system, with a renewal date and terms of service attached. The moment your ability to be believed depends on a company that can revoke it, you don’t own your reputation. You’re leasing it.
There’s an upside buried in this, and it’s worth naming. Once proof is cheap and portable, the people who actually do real work finally get a way to show it that a machine can’t fake in an afternoon. The photographer with the signed original. The contractor with the verified site visit. The founder whose call was signed at the source. In a world drowning in synthetic everything, provable work becomes the scarcest thing there is.
But only if the checkpoint stays a tool instead of turning into a gate.
Because there’s a line here, and it deserves saying plainly. Verifying that a thing is real is not the same as deciding that it’s true. The first is infrastructure. The second is authority. Collapse the two and you get a plagiarism detector that behaves like a referee, a content label that behaves like a censor, and a business model where the companies that made money off your attention now charge you for the privilege of being believed.
For most of the internet’s life, trust was the default and proof was the exception. That’s flipping. Proof becomes the default, and the humans who can’t produce it get filed under noise. The thing worth defending is simple: getting believed shouldn’t require renting permission from the people who profit when you can’t prove yourself.
Get your keys. Sign your work. Make it provable on your terms.
You’ve got this.