Privacy & security

Gennode is designed to handle sensitive data (genomes, scientific datasets), so privacy is a first-class design goal — not an afterthought. The protections below describe the planned data layer and compute model; they are part of the roadmap and not yet live.

How data will stay private (planned)

• Encrypted client-side — data is intended to be encrypted before it ever leaves the owner, so nodes store shards they cannot read.
• Sharded & distributed — no single node will hold a whole file (see the Data layer).
• Access by key — only the owner's keys decrypt. Nothing is sold or profiled.

Compute privacy (roadmap)

• Sealed containers + self-host mode — run jobs 100% locally.
• Confidential computing (GPU TEE) — operators cannot see data even while it is being processed, with cryptographic attestation.

For node operators

• Today, nodes earn for uptime and a CPU benchmark that proves real hardware — they do not yet process user data.
• As the network matures, nodes will run sealed workloads — providing compute/storage without seeing user data.
• Wallet-bound identity + anti-sybil keep the network and rewards fair.

Compliance

GDPR-aware by design: data minimization, consent-based, owner-controlled. As the network matures, sensitive workloads will follow applicable data regulations.

Found a vulnerability? Please disclose responsibly via the security policy on GitHub.