What Happens to Your Evidence When the Capture Vendor Shuts Down?

July 12, 2026

Here is a question worth asking before you trust any service with evidence: if this company disappears next year, does my capture disappear with it?

It is not a hypothetical. Litigation runs for years, regulatory matters run longer, and the average venture-backed startup does not. Even giant vendors kill products and delete the contents on a schedule. This post walks through what actually happened to users’ data in several real shutdowns, what the major capture vendors’ own terms say about your files when you stop paying, and what a vendor-proof capture looks like.

A tale of one team and two shutdowns

In October 2024, Akord, a startup offering encrypted permanent storage, announced it was sunsetting its product: the app, the API, everything. It had onboarded 14,000 users and stored over 2 million files. As of this writing, akord.com no longer loads.

But Akord stored its users’ files on Arweave, a permanent decentralized storage network, and its shutdown announcement could therefore say something remarkable: “any data uploaded to Arweave using Akord, whether public or private, will remain accessible.” Its FAQ went further: “all your data in permanent vaults is on Arweave so will never get deleted.” Users could note their transaction IDs and locate their files on the network directly, with or without Akord. The company died. The data did not.

Now the second act. The same team relaunched as Tusky, building on a different storage network, one based on renewable storage leases rather than permanence. In December 2025, about fourteen months later, Tusky announced it was shutting down too. This time the message was very different: the company said it “cannot currently guarantee users will be able to back up and export their data stored on Tusky” after the cutoff, and its retrieval guide warned that data “will eventually become inaccessible unless recovered.”

Same founders, same product category, two shutdowns a year apart. The variable that decided whether users’ files survived was not the company’s competence or intentions. It was where the bytes lived. Storage that exists only while someone keeps paying and keeps operating dies with the operator. Storage on a permanent network does not.

There is one honest footnote, and it matters for evidence work: Akord’s encrypted private vaults still depended on recovery tooling to find and decrypt them. Bytes that survive are not enough if the tools to read them do not. A truly vendor-proof capture needs both permanent storage and an open, documented format anyone can verify without the vendor. Keep that in mind; we come back to it.

This keeps happening

Storify, 2018. Journalists spent eight years using Storify to preserve social media narratives: breaking news threads, deleted-tweet controversies, primary-source timelines. Adobe acquired it and shut it down; on May 16, 2018, the content ceased to exist, with a few months’ warning to export. Storify’s own cofounder was candid about the logic: “For Adobe, it’s not a product that makes money for them. I understand their decision. Their goal is not to defend journalism.”

MySpace, 2019. Not even a shutdown, just a botched server migration at a still-operating company. MySpace’s statement: “any photos, videos, and audio files you uploaded more than three years ago may no longer be available on or from Myspace,” followed by the suggestion that users “retain your back up copies.” Press coverage put the loss at roughly 50 million songs from 14 million artists, twelve years of uploads.

Google+, 2019. When Google retired its social network, it announced it would “start deleting all content, including Google+ pages, photos and videos” on April 2, 2019. A trillion-dollar company with world-class infrastructure deleted everything, on purpose, because the product no longer fit. Vendor size is not vendor permanence.

Even the Internet Archive has had a hard stretch. In October 2024 a breach exposed 31 million user records and DDoS attacks knocked archive.org offline for days, after which the Wayback Machine ran read-only while systems recovered. On the legal side, the Archive lost the Hachette book-lending appeal in 2024 and settled a record-label lawsuit in 2025 in which reported potential statutory damages ran to $621 million. The Archive survived, and the web is lucky it did. But a single nonprofit under that kind of pressure is not a place to park the only copy of your evidence, a point we cover in depth in our guide to Wayback Machine limitations for evidence.

What the capture vendors’ contracts actually say

The evidence-capture incumbents are established businesses, and nothing here suggests they are going anywhere. But their own terms of service answer the question of what happens to your captures the day you stop paying, and the answers are worth reading before you build a practice on them.

Page Vault’s terms state that after you cancel, your data “will remain on our server for a minimum of seven (7) days,” after which Page Vault “reserves the right (but has no obligation) to remove, delete, or destroy any of the data and content stored in its server.” Seven days is the floor. Getting help with your data after closure is billed at consultant rates.

Pagefreezer’s subscription terms (the February 2022 government-channel version; commercial terms may differ) give you 30 calendar days after termination to request your data back, “subject to applicable fees,” with export pricing listed at $500 for the first 200GB. After the window closes, Pagefreezer “shall have no obligation to maintain or provide any End-User Data” and may “delete all End-User Data without notice or confirmation.”

To be fair to both companies, these are ordinary SaaS terms. That is exactly the point. The standard commercial arrangement for stored data is: it exists while the relationship exists. Evidence has longer obligations than that. The capture you make for a matter today may need to be produced, re-verified, or defended five years from now, past subscription lapses, firm changes, acquisitions, and product sunsets you cannot predict.

The checklist: is your evidence vendor-proof?

Ask four questions of any capture or archiving service, including ours:

1. Where do the bytes live? On the vendor’s servers or cloud account, the data’s lifespan is the vendor’s lifespan (or your subscription’s, whichever ends first). On a permanent decentralized network, the data outlives the account and the company.
2. Can you verify the capture without the vendor? If checking integrity requires logging into the vendor’s portal, the proof dies with the portal. Cryptographic hashes and signatures that anyone can check independently do not, and they anchor the chain of custody besides.
3. Is the format open? A capture in a proprietary format needs the vendor’s software forever. Open formats with multiple independent tools survive, as Akord’s encrypted-vault footnote shows.
4. What do the terms say about termination? Read the data-retention clause before you need it. Seven days is not an evidence-retention policy.

How Permavault answers its own checklist

Permavault was designed so that the answer to “what if you disappear?” is: nothing happens to your evidence.

Captures are stored on Arweave, a permanent decentralized network of roughly 300 independent nodes, funded by a long-term storage endowment rather than our monthly bills; the same property that let Akord’s users keep their files is what your captures inherit, by design. Every file is fingerprinted with cryptographic hashes at capture, and the record is verifiable by anyone without trusting us or logging into anything of ours. Captures use open formats with independent tooling, so reading and verifying them never requires our software. Both halves survive: the evidence and the proof.

Each capture is $4.99. The price is on the page. An optional Certificate of Authenticity, from $9, documents the capture process, and the Legal tier adds a qualified electronic timestamp from Disig a.s., an EU-listed qualified trust service provider, applied to the signed capture manifest, plus an independent Bitcoin-anchored timestamp and a declaration template designed to support authentication under FRE 902(13) and 902(14). Under eIDAS Article 41, a qualified electronic timestamp carries a presumption of the accuracy of its date and time in EU courts.

Vendors shut down. Products sunset. Migrations go wrong. The way to hold evidence through all of it is to make sure the evidence was never hostage to any of it, including to us.

This article is general information, not legal advice. Vendor policies and terms described here reflect published sources as of July 2026 and may change; always review a provider’s current terms directly.