trust & security
Last updated: September 16, 2026
Standalone prediction text is classified in memory and discarded. When Content Gate evaluation collection is enabled, redacted text is stored privately for up to 90 days to measure classification quality. Automated redaction may leave personal information, so these records remain access-restricted. This page documents how we handle data, where we process it, and the controls we run.
/v1/predict is classified in memory and not persisted to any database, log, error report, backup, or support system.indeterminate, never allow.Customer data is processed and stored in the United States. Databases (DynamoDB), transactional email (SES), and storage run on Amazon Web Services in the us-east-1 (Northern Virginia) region; application compute runs in a US region.
Content Gate raises the cost of automated form spam and assesses the exact text submitted to a customer's backend. It does not prove humanity, replace bot or DDoS protection, or protect a form unless the customer verifies the response on its backend before performing side effects. The widget uses no fingerprinting, advertising identifiers, or cross-site cookies. Submitted text and challenge material are not retained or logged; short-lived content-free verification results support replay safety and idempotent retries.
The third parties that process customer data on our behalf:
| Subprocessor | Purpose | Data | Location |
|---|---|---|---|
| Amazon Web Services | Hosting, database, transactional email | Account & request metadata, emails | United States |
| Stripe | Payment processing & billing | Billing details | United States |
| Google Analytics | Opt-in website analytics (marketing site only) | Aggregate, pseudonymous usage metrics | United States |
| Fontshare | Web-font delivery (marketing site) | None beyond standard request logs | — |
None of these subprocessors receives the content you submit to /v1/predict.
Siftfy acts as a data processor for the limited request metadata it stores on your behalf and as an independent controller for account and billing data; processing takes place in the United States. If your organization requires a separate Data Processing Agreement, email hi@siftfy.io and we will work with you to put appropriate terms in place, including the Standard Contractual Clauses where they are required for transfers from the EEA, the UK, or Switzerland.
A full retention schedule is in our Privacy Policy. Standalone prediction content is not retained; enabled Content Gate evaluation text is logically retained for up to 90 days; logs roll off within 90 days; revoked keys within 90 days; magic-link tokens within 15 minutes; billing records are kept as legally required; and account data is deleted on request.
Found a security issue? Email hi@siftfy.io. We acknowledge reports promptly and will work with you on a responsible disclosure timeline.
For standalone /v1/predict requests, no. Enabled Content Gate evaluation separately retains redacted text in private storage for up to 90 days. The text you send to /v1/predict is classified in memory and is not written to any database, log line, error report, backup, or support tool. Once Siftfy returns the score, the content is gone from our systems.
No. Our application logs and our validation/error responses are built to carry only metadata — request ID, status, timing, detected language, the spam probability, and account identifiers. They do not include the submitted text. We do not run a third-party error-capture or session-replay tool that would ingest request bodies.
In the United States. Siftfy's databases, transactional email, and storage run on Amazon Web Services in the us-east-1 (Northern Virginia) region, and our application compute runs in a US region.
If your organization requires a Data Processing Agreement, email hi@siftfy.io and we'll work with you to put appropriate terms in place, including Standard Contractual Clauses where they're required for transfers out of the EEA, UK, or Switzerland. This Trust & Security page documents our security controls, subprocessors, data residency, and retention.
We retain account email and account metadata for the life of the account; request metadata (timestamps, status, duration, language, spam probability, tenant and key identifiers) and usage counts to bill and operate; server and application logs for up to 90 days; revoked API keys for 90 days; magic-link sign-in tokens for 15 minutes; and billing records as required by tax law (typically 7 years). Separately, enabled Content Gate evaluation collection retains restricted redacted text for up to 90 days.
Standalone prediction content is not retained or used for training. When Content Gate evaluation collection is enabled, we retain redacted submissions for up to 90 days to evaluate quality, with restricted access and human labels. Automated redaction does not guarantee anonymity; these records are not used for automatic training. If you opt to submit a ham/spam correction to /v1/feedback, we use that text only to compute a salted one-way fingerprint; we store the fingerprint and label to improve accuracy, never the readable content. Marketing-website analytics are first-party and opt-in and never see API content.
On a deletion request we remove your account email, API keys, passkeys, usage records, and the feedback fingerprints attributed to your account; backups containing that data roll off within about 30 days. The only data we keep is billing and tax records we are legally required to retain.
Questions about security or data handling? Email hi@siftfy.io.