Introducing KYCO: Know Your Content and the Future of Digital Trust

The problem is bigger than fake images

Back in 2020, I presented an idea for a synthetic media identity system to the European Advertising Standards Authority. We were building virtual humans and conversational systems, and I could see a growing problem: people would increasingly struggle to distinguish a real person from a synthetic character or a coordinated network of accounts.

My concern was not simply whether AI had helped create an image. It was whether people could understand who stood behind the content and what their motivation might be.

A fictional character can be entertaining. A brand-owned avatar can be useful. The difficulty begins when their identity and purpose are concealed.

A crisis of authenticity

Generative tools make it easier to create convincing voices, faces, video and conversation. That brings extraordinary creative opportunities, but it also lowers the effort needed to impersonate people and manufacture apparent legitimacy.

For a smaller business, trust is already hard to earn. If fraudulent advertisers can surround a fake offer with synthetic testimonials, invented people and plausible websites, legitimate companies face a more suspicious audience too.

The same challenge affects public debate. A wave of apparently independent opinions may actually come from automated accounts under common control. Without better signals about origin, it becomes harder to judge what we are seeing.

Three questions we need to separate

I think a useful system should help answer three distinct questions.

Who is responsible for publishing this? How was the content created or changed? What evidence supports the claim it makes?

These questions are connected, but they are not interchangeable. A verified human can publish something false. An AI-generated image can be clearly labelled and entirely legitimate. A provenance record can document an editing history without proving that the scene depicted is true.

What KYCO proposes

KYCO stands for Know Your Content. It is my proposal for combining publisher verification, attribution and synthetic media identification in a more coherent system.

The inspiration comes partly from the way commerce systems establish who is participating and which actions they are authorised to take. I think similar principles could help content platforms distinguish an accountable publisher from an unverified impersonation.

Imagine a business being independently verified before receiving a meaningful trust signal on its advertising. Then imagine content carrying evidence that it came from an authorised publisher, with a way to revoke that credential when it is misused.

This is a proposal for exploration, rather than a claim that a complete KYCO product has already been deployed. Its usefulness would depend on implementation, adoption and independent scrutiny.

Privacy must be part of the design

Verification should not require every platform to hold a person's full identity documents.

Privacy-preserving credentials could allow a service to check a necessary attribute while receiving less personal information. Zero-knowledge techniques are one possible component, but they do not remove every difficult question about issuance, misuse, account recovery or accountability.

The design would also need to protect legitimate anonymity. Whistleblowing, political dissent and personal safety create real reasons for people to speak without publishing their legal identity.

Build on provenance, without confusing it with truth

C2PA provides standards for recording media provenance. Watermarking systems such as Google DeepMind's SynthID offer another way to help identify certain AI-generated material.

These approaches address different parts of the problem. A wider trust system would need to connect useful creation signals with publisher accountability, clear presentation and routes for challenge and redress.

No badge should imply that everything a publisher says is true. No missing badge should automatically mean that someone is lying.

Trust is an infrastructure question

I have championed synthetic media for years. I want its creative and commercial potential to flourish. That requires an environment in which people can make informed choices about what they see and whom they trust.

KYCO is an invitation to think across identity, content, privacy and platform incentives together. Better detection can help, but the longer-term opportunity is to make trustworthy participation easier to establish in the first place.