The Verification Economy: Why the Next Fortunes Won’t Be Made Creating Things
AI made everything easier to make. It also made everything harder to believe.In early 2024, a finance employee at the engineering firm Arup joined a video call. His CFO was on it, and so were several colleagues. He recognized their faces and voices, the familiar grid of a corporate Tuesday. They…
AI made everything easier to make. It also made everything harder to believe.In early 2024, a finance employee at the engineering firm Arup joined a video call. His CFO was on it, and so were several colleagues. He recognized their faces and voices, the familiar grid of a corporate Tuesday. They discussed a confidential transaction, and over the following days he made fifteen transfers totalling about $25 million.Every person on that call except him was a deepfake.The indicators people have used as proof that something is ‘real’ their whole lives have stopped proving anything. A face was a person; it meant a body in a room somewhere. A photograph used to mean that light had, at some point, hit a lens. Banks, newsrooms, HR departments, your own phone, suddenly and intensely are suspect.Ask what AI can generate and the answer is now boring: everything. Infinite images, infinite video, infinite voices, infinite text, infinite music. The world (and its stockholders) is transfixed by the supply side: look how much we can make, look how fast, look how cheaply. Perhaps a more interesting question is: is what AI makes unreliable?In 1993, The New Yorker ran what became the most reprinted cartoon in its history. Two dogs at a computer, “On the Internet, nobody knows you’re a dog.” For thirty years, that read as a gag about anonymity, the internet as a costume party. What if that cartoon from 30 years ago was a warning? Suddenly, the costume extends to the thief who looks and sounds exactly like your boss. A new house of mirrors, distorted reality where anyone can be anyone and anything can be anything.AI is creating a verification economy.Economics does a predictable thing when something becomes abundant, and it’s older than computing. When something becomes abundant, something else becomes scarce. Value moves to that scarcity. If artificial intelligence makes creation infinite, creation cannot be the valuable thing anymore. Lasting value can’t come from a tap.What if the wealth of the AI era doesn’t come from data centres or fintech or even chips? What if it comes from proof?Proof that the person asking you to wire money is your actual boss. Proof that the remote job applicant exists. Proof that the photo came from a camera on this day at this time rather than ChatGPT five seconds ago. Proof that the voice on the phone saying “Mom, it’s me, I’m in trouble” belongs to an actual kid. (That scam stopped being hypothetical, so if you haven’t warned your parents yet, here’s your friendly reminder.) Proof that the thing in front of you has some reliable connection to reality.More generations mean more fraud, more impersonation, more uncertainty, which means permanent, civilization-wide spending on certainty.It begs the question: what actually can’t be faked?I went hunting for it. What is the thing that can’t be faked, and let’s build the future on it. Let’s invest in it. A better detector? Maybe. A heroic AI trained to catch the others? Or cryptography, an object with certainty sealed inside?A face can be faked, and so can a voice. A signature, the mark civilization has been putting at the bottom of documents for thousands of years, might be the easiest item on the list. An ID can be forged, or built around a synthetic person with a generated employment history, generated references, and a smiling headshot of someone who was never born.Even the cryptographic key, the hunt’s most promising ending, can be stolen, or used correctly by exactly the wrong hands.In July 2025, OpenAI’s Sam Altman told a Federal Reserve conference that finance was facing a “significant impending fraud crisis.” AI has “fully defeated” most of the ways people currently authenticate. The voiceprint was his lead example of what banks still madly accept. According to Altman, one survivor remained in the password. A secret, something you know instead of something you are.But passwords get leaked all the time.Which is why the technical story of the decade is the passkey, and its trick is narrow and real. It never leaves your devices, so control of a credential gets established without the secret ever being transmitted. You know it matters because Apple, Google, and Microsoft jointly committed to the standard. When the three biggest rivals on earth link arms? I pay attention.But passkeys are all about the right credential approving the action. They can’t enforce that the human holding it wasn’t deceived, threatened, or confused about what they were approving. A passkey proves control, not truth. You know those 6-digit codes you get on your phone to approve a log-in? A senior in the midst of a phone scam can still be convinced to give them over in real-time.The same promise is arriving for pictures. A coalition called the C2PA built an open standard that lets a camera cryptographically sign a photo at the moment of capture. Leica shipped the first camera with it built in, and the standard is spreading toward phones. A lovely idea, but a camera will faithfully sign a photograph of a staged scene. Or a real image can carry a falsified caption. An honest and real video recording can be cut and edited at the exact second that reverses its meaning. Provenance isn’t truth either.The hunt continues. Authentication proves control. Provenance proves history. How about accountability?In July 2024, a Ferrari executive got WhatsApp messages from CEO Benedetto Vigna about a confidential acquisition. It was from a wrong number, but when the call came, it had the right voice, a spot-on southern Italian accent.The executive caught something mechanical in the intonation and said, “Sorry, Benedetto, but I need to identify you,” and asked what book he had recommended to him. The line went dead. What stopped the fraud was a fact that lived inside a real relationship. We can see this idea having a ripple effect already in today’s economy. Trusted brands are growing more valuable, not less.If every consequential interaction now carries a new question bolted to it — was this person real, was this image real, was this transaction real — then whoever owns the trust layer, the standards and credentials everything must pass through, gets paid continuously. Not once. Continuously. That’s the finest business model humans have ever invented: the tollbooth.The market is obsessed with generating things. The next decade will belong to the people who can prove them.J.D. Macpherson is a Canadian journalist and award-winning author of the narrative nonfiction book Human Again in the AI Age.This story is published under the Generative AI publication. Connect with us on LinkedIn and follow Zeniteq to stay in the loop with the latest AI stories. Let’s shape the future of AI together!The Verification Economy: Why the Next Fortunes Won’t Be Made Creating Things was originally published in Generative AI on Medium, where people are continuing the conversation by highlighting and responding to this story.Source: Generative AI Pub — Published — Category: Image AI