TRUST & SECURITY
Security & Trust at Press Run
Press Run holds the keys to your social accounts and publishes in your name. That only works if we handle your credentials and your content more carefully than you would yourself. This page says, in plain language, what we actually do today — and, further down, what we don't do yet.
How we protect your account
Your password is never stored. Passwords are hashed with Argon2id, the current industry-standard hashing algorithm, before they touch our database. There is no copy of your password anywhere in Press Run to steal.
Two-factor authentication is built in. Every account can turn on authenticator-app 2FA (TOTP) with one-time recovery codes. Organizations on the Enterprise tier can require 2FA for every user — and lock that requirement so nobody below corporate can switch it off.
Your connected-account tokens are sealed. When you connect a Facebook, Instagram, or other platform account, the access tokens we receive are encrypted at the application level (Fernet sealing) before storage, and they are never exposed through our API — not even back to you. The same sealing protects our most sensitive stored data, including Brand Persona reference embeddings.
Sign-in tokens are digest-only and short-lived. Email verification links expire in 24 hours; password-reset links expire in 30 minutes, work once, and signing in again revokes every other session.
We keep the receipts. Sign-in and account-security events are written to an audit log, and every content approval records who decided it and when.
How we protect your data
Encryption in transit and at rest. All traffic between you, Press Run, and the platforms runs over TLS. Data at rest sits on infrastructure with provider-managed disk encryption, and — as described above — credentials and tokens get a second, application-level layer of encryption on top of that.
Your data is yours, in a database of our own. Press Run runs on its own dedicated database, its own login system, and its own workers — it is not a shared multi-app account system. One customer's data is scoped to their organization by design: users see only the part of the organization tree their role covers, enforced on every read, not just hidden in the interface.
Backups are verified, not assumed. Our production database is backed up before every significant change, with checksum (SHA-256) verification of each backup file, and full database backups are copied weekly to separate cloud storage. Our restore process has been tested end-to-end against those backups — most recently verified on November 17, 2026.
Governance is a security feature. Enterprise organizations get locked policies (channels corporate bans cannot be enabled below), approval routing (nothing publishes without the approver corporate chose), and a per-location kill switch that stops publishing immediately. An AI draft that cannot cite a source for its claims is rejected by the system rather than published.
Who else touches your data
We use a short list of infrastructure and service providers — hosting, payments, email delivery, AI generation, and the social platforms you connect. The full list, with what each one processes, is published here: Subprocessors → We will update that page before we add a provider that processes customer content.
Reporting a security problem
Found something? Email [email protected]. We read every report, we will acknowledge yours within two business days, and we will not take action against good-faith researchers who report privately and give us reasonable time to fix an issue before disclosing it.
What we don't have yet — stated plainly
Credible vendors say this part out loud, so we will:
- We are not SOC 2 certified yet. An independent SOC 2 audit is on our 2027 roadmap, after our first enterprise pilots. Today, the practices described on this page are real and in production — they simply have not yet been audited by an outside CPA firm, and we won't claim otherwise.
- Single sign-on (SSO/SAML) is in development for the Enterprise tier. It is not available today. Enterprise customers currently sign in with Press Run accounts protected by the 2FA described above, which organizations can mandate by policy.
- We have not yet completed an independent penetration test. Internal security testing is part of every release; a third-party test is scheduled as part of the same 2027 compliance program.
- We do not offer a contractual uptime SLA yet. Our service status and uptime record are published on our status page. Until we have the track record to back one, we won't sell you a guarantee we can't prove.
If your security team needs more than this page — a completed questionnaire, a call with the founder, our subprocessor list in writing — ask. We answer security questionnaires from our published facts, and we'd rather tell you "not yet" than dress up an answer.
Press Run is built and operated by Vajgrt Labs.