Security and privacy
How we protect your account, your ad account connections and your API keys.
Workspace isolation
Every campaign, piece of content, task, file and connection belongs to one workspace. Your membership and role are verified on the server for every request; one workspace's data can't be seen from another. Permissions are role-based — see Roles and permissions.
Encryption
- Meta and Google access tokens and BYOK API keys are stored encrypted with AES-256-GCM.
- Encrypted keys are never sent to the browser; the UI only shows a masked version.
- Keys are never written to logs.
- All traffic travels over HTTPS.
Ad account connections
Meta and Google connections use the official OAuth flow; we never see your password. The connection request is protected by a signed, single-use verification code, so nobody else can attach your account to their workspace. You can revoke access at any time from Adflap or directly in your Meta / Google settings.
You always control spend
Campaigns created by Adflap start paused, and every AI assistant action that spends money waits for your approval.
Account security
- Passwords are at least 8 characters and stored one-way hashed.
- Sign-in, sign-up and password reset attempts are rate-limited.
- Session cookies are
HttpOnly; sessions last 30 days. - Under Settings → Profile → Active sessions you can sign out devices you don't recognize, and sign out all other sessions when changing your password.
File and link safety
An uploaded file's real type is checked from its contents; corrupted or mismatched files are rejected. Pages analyzed for one-click campaigns and files imported from a URL are fetched by a protected downloader that blocks access to internal network addresses.
AI and your data
For AI actions, only the data needed for the job (brand profile, campaign summary, the text you asked about) is sent to the model provider. If you use your own key (BYOK), data goes directly to the provider you chose.
Activity log
Important actions — publishing campaigns, AI approvals, adding or removing keys, content changes — are written to the activity log along with who did them.
Reporting a security issue
If you think you've found a security vulnerability, please write to [email protected]. We review these reports with priority.