mandatiq

Security

Where your data lives, who sees it, how to report issues

The short version of how mandatiq handles your data. The full picture lives in the Privacy Policy; this page is the operational summary an engineering team can scan in a minute.

Where your data lives

Customer scan data — URLs, HTML snippets, screenshots, AI fix suggestions — stays in the EU. The non-EU sub-processors above are bound by Standard Contractual Clauses (2021/914) and the EU–US Data Privacy Framework where applicable.

Who has access

What we send to AI providers

AI fix suggestions are generated by the Claude API (Anthropic PBC). For each violation we send: the WCAG criterion ID, the failing rule (e.g. color-contrast), and a short HTML snippet of the failing element. We do not send full page HTML, screenshots, customer names, or any account metadata. Anthropic does not train on data sent through their API.

Encryption

Abuse + bot defence

The free anonymous scanner is gated by Cloudflare Turnstile (privacy-respecting CAPTCHA, no third-party cookies), per-IP and per-domain daily limits, and a global daily budget cap to keep the system available during traffic spikes or scripted abuse.

Reporting a security issue

Email security@mandatiq.com with a description and a way to reach you for follow-up. We aim to acknowledge within 72 hours and to ship a fix or a mitigation as quickly as the severity allows. Please do not file public GitHub issues for unpatched vulnerabilities. Researchers should also read our SECURITY.md for the in-scope/out-of-scope list, embargo policy, and known accepted risks.

Questions about the data we collect, retention windows, or your GDPR rights as a data subject belong on the Privacy Policy — that's the legally-binding source.