The notarization/anchoring graveyard: lessons, not schadenfreude

Several blockchain-based document-notarization and data-anchoring projects from the 2015-2020 era shut down, pivoted, or were acquired and quietly folded into something else. This page is a respectful look at what changed and why — the point is to be honest about what a longevity-focused evidence system needs to survive, not to mock anyone's shipped work.

Factom

Factom built a blockchain-anchoring layer for arbitrary data integrity (its Entry Credit model recorded hashes into its own chain, itself periodically anchored into Bitcoin). The project wound down its original foundation-and-token model in the early 2020s; the underlying technology and team continued in different forms (Factom Protocol / later work under different branding). Widely reported at the time as a token-economics and go-to-market sustainability problem more than a cryptographic one — the anchoring mechanism itself wasn't shown to be broken.

Stampd / similar document-timestamping startups

Several smaller blockchain-timestamping products from the same era (cheap, simple "hash your document into a blockchain" tools) did not survive as standalone businesses — commonly because the value proposition (a timestamp proof) was thin on its own without the surrounding identity, authorization, and admissibility context a real dispute or audit actually needs.

Lesson 1 — a timestamp alone is not evidence. "This hash existed at time T" answers one narrow question. It says nothing about WHO produced the data, WHETHER they were authorized to, or whether the surrounding chain of custody holds up under FRE 901/902-style scrutiny. ActaSeal's dispute packets bundle acquisition/chain-of-custody/ authentication documents alongside the hash chain specifically because a bare timestamp was never the actual product a dispute needs.
Lesson 2 — dependency on a single external chain/service is a longevity risk. A notarization product whose entire value depends on one blockchain's continued operation (or one company's continued operation) inherits that chain's or company's own sustainability risk. This repository's own multi-anchor posture (RFC 3161 pool + transparency log, never a single point of witness) — built as a direct response to this exact failure mode.
Lesson 3 — token/business-model sustainability is a separate risk from cryptographic soundness. Several of these projects' underlying cryptography was never shown to be broken; what failed was the business wrapped around it. This system is designed so that verification survives the operating company ceasing to exist: every packet embeds its own trust root (the signing key is carried in the manifest, so no live service or key registry is ever consulted), and the standalone verifier (verify.py, this repository) has zero dependency on any company's continued operation.

Why identity + authorization + admissibility ≠ a notary chain

A blockchain notary answers "did this hash exist at time T." ActaSeal answers a different, larger question: "was this specific agent authorized, under this specific policy, with this specific evidence, to take this specific money-touching action — and can that be proven independently later." The timestamp/anchor machinery is one input into that larger evidentiary record, not the whole product.

← Back