Privacy Policy

What we collect

• Account basics. When you sign in with GitHub or Discord, we store your username, account ID, and avatar URL. Nothing else from those platforms.
• Machine telemetry. If you run the Hadal worker, we store the machine identifier, GPU name, VRAM size, utilization and temperature readings, and contributed hours. This is what powers the public network stats.
• Job results. Outputs of research jobs your machine executes, attributed to a random worker ID.

What we never collect

• No email addresses (unless you later choose to provide one).
• No personal files. The worker only touches job data assigned by the network — it never reads your documents, photos, or browsing history.
• No cookies for advertising, ever. The single functional cookie/localStorage entry is your own login session.
• No analytics or third-party trackers.

How your data is used

• Usernames may appear on public leaderboards and contributor walls (that is the point of community credit).
• Aggregate statistics (total workers, total hours) are public. Individual machine telemetry is visible only to you on your dashboard.
• We do not sell or share data with anyone. There is no revenue model that involves your data.

Storage & security

Data lives in an encrypted managed Postgres database (Neon, US region). Sessions are 256-bit tokens valid 30 days. Worker keys are stored hashed where possible and can be revoked individually at any time from your dashboard.

Your controls

• Delete a key — unlinks that machine instantly.
• Sign out — clears your local session.
• Account deletion — contact us (Discord or GitHub) and we remove your account, machines, and stats. Open-source means you can also verify everything we claim by reading the code.

Children

Hadal Research is not directed at children under 13 (16 in the EEA), and we do not knowingly collect their data.

Changes

If this policy changes materially, we will announce it on the site before it takes effect.