Research paper published: Toward Decentralized Trust and Verifiable Content on the Web

Apr 1, 2026·
Jason Grey
· 2 min read
blog

We’re pleased to share the publication of Toward Decentralized Trust and Verifiable Content on the Web, the paper that lays out the technical foundation for the HTMLTrust project.

Abstract

We propose a decentralized, standards-aligned framework for embedding cryptographic trust directly into HTML content. Using a new <signed-section> element, content creators and publishing platforms can sign semantically meaningful regions of web pages and include identity-linked metadata in-band. Signatures are validated using public key infrastructure (PKI) such as DIDs, and can be enhanced with third-party endorsements submitted to optional, federated trust directories. We introduce a simple canonicalization method for content normalization and outline how browsers and CMS systems can support user-configured web-of-trust policies for live content validation. Unlike blockchain-based or DRM-centric systems, our approach is lightweight, browser-compatible, and web-native — designed to scale across publishing workflows, civic media, and knowledge networks.

Key contributions

  1. A proposed <signed-section> HTML element for encapsulating signed regions of a page
  2. A canonicalization algorithm for consistent content normalization
  3. A trust model supporting both direct signature verification and third-party endorsements
  4. Integration paths for browsers and content management systems

Reference implementations

Alongside the paper, we’ve published reference implementations:

  • Trust Directory Server — Node.js + MongoDB API for author management, content signing, verification, and reputation tracking
  • Browser Extension — Chrome extension for client-side signature validation
  • CMS Plugins — WordPress plugin and Hugo integration for server-side content signing
  • Canonicalization Libraries — JavaScript, Go, and PHP implementations, all passing the same conformance suite

Read the paper

The paper and its LaTeX source are available in the htmltrust-spec repository.

We welcome feedback and contributions. Visit the GitHub repositories to get involved, or reach out at jason@jason-grey.com.

Paper Specification Announcement
Authors
Jason Grey
Author of the HTMLTrust spec
Jason has been building for the web since the early 1990s, with decades of work across digital and interactive media. Since the early 2000s he has worked on AI, search, generated content, and the semantic web. Recent work includes running jobs across Common Crawl’s entire 10PB corpus and building search-and-context infrastructure at Kagi — domains where the trustworthiness of web content is a daily, practical concern. HTMLTrust is a response to that gap.