Synthetic Reality and the Erosion of Evidence.
The danger of synthetic media is not that individuals will be deceived. The danger is that the category of evidence itself becomes optional.
A society can survive a lie. It cannot survive the loss of the category lie.
The collapse of cost
For most of recorded history, a photograph, a recording, a video, a signed document, all carried evidentiary weight. Not because any of them were perfect, but because forging any of them at scale was expensive, slow, and traceable. The cost of fabrication was itself a kind of public trust.
That cost has collapsed. The marginal price of a convincing synthetic image of a real person, a synthetic voice in their cadence, a synthetic video of them saying something they never said, is now negligible. The infrastructure is consumer grade. The skill required is roughly the skill required to use a phone.
This is the part of the story that journalists, courts, election officials, intelligence agencies, and the private security industry already understand. It is not yet the part of the story that the public, on average, has been told plainly.
Why the harm is structural, not anecdotal
It is tempting to focus on individual fakes, the dramatic ones, the political ones, the celebrities. That framing misses the larger phenomenon.
The deeper harm is not that any specific fabrication will deceive a specific audience. The deeper harm is the erosion of any default trust in evidence at all. When everything can be fabricated, nothing has to be denied. The liar wins not by being believed, but by making belief itself a matter of taste. Once that threshold is crossed in a population, every authentic recording of a real abuse, a real statement, a real event, becomes contestable on demand by anyone who finds it inconvenient.
This is not a new dynamic in human history. It is, however, the first time the means of producing it are cheap, ambient, and improving on a quarterly cycle.
The liar wins not by being believed, but by making belief itself a matter of taste.
Shared reality is a public utility
Democracy presumes a shared evidentiary baseline. Courts presume one. Journalism presumes one. Public health presumes one. So does the most ordinary act of citizenship, which is the act of one person, talking to another, about what is happening.
When that baseline is corroded, the institutions that depend on it do not fail loudly. They fail by inches. Juries become less persuadable. Officials become less accountable. Witnesses become less credible. The public square, which is already noisy, becomes one in which no one can be confident which voices in it are coming from human beings.
Maintaining shared reality is, in the present technological era, a civic infrastructure problem. It has to be funded, defended, and built, the way water systems and roads and elections are funded, defended, and built. There is no private market that will provide it without being required to.
What a baseline policy looks like
The minimum policy floor is well understood by the technologists who would implement it. Provenance metadata embedded at the moment of generation. Visible labeling on platforms that distribute synthetic political and impersonating content. Criminal liability for the malicious creation or distribution of unlabeled synthetic media depicting real individuals. Public education funding so that ordinary citizens know what the labels mean.
None of this is novel. None of it is technically out of reach. None of it is being adopted at the speed of the problem.
A public that cannot tell what is real cannot govern itself. The remedy is older than the technology. It is overdue.
