Security

Security you can check for yourself

Our guarantee is not “trust us”. Records are built so that an outsider can detect when they have been changed, including by us.

Threat model

What each attack runs into

Scenario	What happens
An attacker edits a score or certificate in the database	Verification reads from the chain, so the database edit changes nothing.
A video or document is swapped on the server	Its hash no longer matches the published root, and the integrity check reports it.
One blockchain node is compromised and lies	Agreed reads need matching answers from other nodes, so the false answer is refused and logged.
One validator goes down	The other three keep producing blocks. The failed node catches up when it returns.
Two validators go down	The chain pauses by design rather than risk a split. Writes are refused and reads continue.
Someone rewrites the whole chain	Block hashes no longer match the roots already timestamped through OpenTimestamps.
Someone tries to publish under a creator's name	For MetaMask creators the signature fails without their private key, which we never hold.

Controls

Platform-managed wallets are stored with AES-256-GCM.

Google OpenID Connect with PKCE, or Sign-In with Ethereum.

Encrypted video segments with keys checked against the license on every request.

Password-protected archives covering database, chain and media.

Atomic balance updates, single-use approvals and a hash-sealed ledger.

Certificates and ledger events can be inspected without an account.

Being precise

A security claim is only useful if its limits are clear.

It does not stop content from being copied by someone with legitimate access.
It detects changes but does not restore data. That is what backups are for.
It cannot prove that information was correct when it was first entered.
Learners using platform-managed wallets rely on the platform to protect those keys.

See how ChainLMS fits your organization, or start learning today.