Citizens have a right to know whether what they are seeing, hearing, or reading was produced by a human being or by a machine.
A public that cannot distinguish a real recording from a fabricated one cannot hold anyone accountable for anything. The right to know what is real is foundational to a functioning democracy. Labeling does not solve every harm, but its absence guarantees that fraud, harassment, and political manipulation scale faster than any institution can respond.
A democracy depends on a shared evidentiary baseline. When that baseline collapses, citizens lose the ability to distinguish testimony from fabrication, journalism from manipulation, and political speech from synthetic forgery.
That collapse is no longer hypothetical. Voice clones have been used in political robocalls. Fabricated images have been circulated as evidence of events that did not occur. Synthetic video has been deployed to impersonate public officials and private citizens alike. The technology is available, cheap, and improving.
A clear, enforceable requirement that synthetic and substantially AI-modified media be labeled at the point of creation, carried through distribution, and visible to the end viewer.
Not a ban on synthetic media. Not a restriction on creative use. A label. The same principle that requires nutrition information on food, disclosure on advertising, and provenance on currency.
Machine-readable provenance metadata embedded at the moment of generation. Visible on-screen indicators on distribution platforms. Penalties for the deliberate removal of such labels. Liability for platforms that knowingly distribute unlabeled synthetic political or impersonating content.
None of this is technically novel. The standards exist. The capability exists. What is missing is the requirement.
- 01.Mandatory provenance metadata embedded by generators of synthetic image, audio, and video content.
- 02.Visible labeling on distribution platforms for political, journalistic, and impersonating content.
- 03.Criminal liability for the malicious creation or distribution of unlabeled synthetic media depicting real individuals.
- 04.Platform obligations to detect, label, or remove unlabeled synthetic content in election windows.
- 05.Public education funding so citizens know what the labels mean and how to verify provenance.
- —Every recorded statement by a public figure will be deniable.
- —Every piece of documentary evidence will be contestable.
- —Authentic footage of real abuses will be dismissed as fakes.
- —Trust in shared reality, already strained, will fail in the moment it is most needed.
- Objection
- Answer
- Objection
- Answer
- Objection
- Answer
"Labels will be ignored."
Some will. Most will not. Nutrition labels, surgeon general warnings, and content ratings are imperfect and still indispensable. Imperfect signal is better than no signal.
"This is a free-speech problem."
Disclosure is not censorship. Requiring that synthetic content be identified as synthetic does not restrict what may be said. It restricts the right to pretend a machine is a human.
"Bad actors will simply strip the labels."
Then stripping the label becomes the prosecutable act. The point of the rule is to make the deception itself unlawful.
A public that cannot tell what is real cannot govern itself. The remedy is not novel. It is overdue.
Human Future Framework · Information Integrity
- → Generate a call/email script tailored to this demand, then phone or write your officials.
- → Find your representatives and contact them about this demand.
- → Sign the Human Future Pledge to add your name (or anonymity) to the public record.
- and log the share below.