For most of photography's history, a photo was treated as proof almost by default. If you had the picture, you had the evidence. That assumption is breaking down — not because fake photos are new, but because anyone can now plausibly claim a real photo is fake.
The liar's dividend
Researchers have a name for this: the liar's dividend. As deepfakes and AI-generated images become common knowledge, people gain a new defence against real evidence — simply claim it was fabricated. It's already shown up in real cases. Tesla's legal team argued that past public statements about self-driving safety couldn't be used against the company because they might be deepfakes. A defendant in a Capitol riot case argued the video evidence against him could be AI-generated. Neither claim needed to be true to be useful — raising the doubt was often enough.
Why this matters beyond the courtroom
You don't need a lawsuit for this to bite. Consider a few everyday situations:
- A tradesperson's before/after photos get challenged in a customer dispute — "how do I know that's actually from my property, on that day?"
- An insurance claim photo gets questioned by an adjuster looking for any reason to delay or deny.
- A workplace incident report gets picked apart during an investigation, with the photographic evidence itself becoming the point of dispute.
In each case, the person relying on the photo now has to answer a question that never used to come up: how do we know this is real?
What actually answers that question
A photo alone doesn't answer it — not anymore. What does is the data around the photo, captured at the same moment, that can't be edited after the fact: exactly when it was taken, exactly where, and a way to prove the file hasn't been altered since. That's a timestamp, a GPS coordinate, and an integrity hash — not glamorous, but it's the difference between "trust me" and "check for yourself."
What Preserver does about it
Every capture in Preserver — photo, video, voice note or text — is stamped with time, date and GPS the instant it's recorded, and protected with a SHA-256 integrity hash. None of that is editable after the fact. It doesn't make a fake image real, and it doesn't stop someone from claiming your photo is fake — but it gives you something to point to besides your word.
The best evidence is the kind that doesn't need you to vouch for it. The timestamp and hash speak for themselves.
The short version
Anyone can generate a convincing fake photo now, and anyone can claim a real one is fake. The way through that isn't a better photo — it's proof that travels with it. That's the gap Preserver is built to close.
