Demand that AI-generated images, videos, and voices remain watermarked and restricted until the user verifies their identity before download, posting, export, or public use.
signatures
Three foundational requirements to restore trust in digital media.
All AI-generated images, videos, and voices should be watermarked, restricted, and cross-checkable by default. Creation remains broadly available — but outputs stay locked until identity is verified.
Users may create normally, but before downloading, posting, exporting, commercializing, or removing restrictions, they must complete identity verification.
The AI provider should retain the original generated asset and tie any public use or unrestricted release to the verified person or entity responsible.
AI-generated deepfakes are being weaponized at scale. Without watermarking by default and identity verification before release, anyone can create and distribute photorealistic fake images, videos, and voice clones anonymously — destroying reputations, manipulating elections, and enabling fraud.
AI tools are being used to generate explicit fake images of real people without their knowledge or consent, causing severe psychological harm.
Deepfake videos of political figures saying things they never said threaten democratic processes worldwide.
AI-generated voice clones, video deepfakes, and synthetic recordings are being used to impersonate executives, authorize fraudulent transfers, and manipulate legal proceedings.
As deepfakes become indistinguishable from reality, society loses the ability to trust any digital media — the "liar's dividend."
We, the undersigned, call upon lawmakers, regulators, and technology companies worldwide to implement mandatory watermarking and restrictions by default, KYC/AML before public use or unrestricted release, and original-asset storage for AI-generated images, videos, and voices.
Artificial intelligence has made it increasingly easy to generate photorealistic images, videos, and voice clones of real people doing and saying things they never did. These "deepfakes" can be created in seconds, distributed globally in minutes, and cause irreparable harm to individuals and institutions alike. The technology that enables this operates with zero accountability — no watermarking by default, no identity verification before release, no recourse for victims.
The right to control one's own likeness is fundamental. The right to trust what we see and hear is essential to a functioning society. Without action, we face a future where any image, video, or voice recording could be fabricated, where truth becomes indistinguishable from fiction, and where the most vulnerable members of society bear the greatest cost.
We do not seek to ban AI technology. We seek to ensure it is used responsibly, with accountability, and with respect for the rights and dignity of every person. Watermarking by default and identity verification before release are proven, implementable solutions that balance innovation with protection.
The time for voluntary self-regulation has passed. We demand legislative action now.
Your verified signature matters. Let's make a difference together.
We'll send a verification code to confirm your email.