How it works

An independent watch on your public presence, in four honest parts.

1. Connect public profiles

Paste the public URLs or handles for the profiles you want watched — X/Twitter, LinkedIn, Instagram, Facebook or Threads. We never ask for your password; the MVP only reads what is already publicly visible. OAuth-based, platform-verified connection is coming soon.

2. Continuous watch

Guardian AI builds a behavioural baseline for each profile: how often it posts, the tone and vocabulary, the hours it's active, and the links it shares. We then watch for drift — sudden floods of posts, mass replies, off-brand political or financial spam, unusual language, activity at odd hours, and bio or link changes.

3. Instant alert on anomalies

When activity crosses an anomaly threshold, we raise an alert immediately: in-app and by email (SMS is coming soon). Each alert explains what changed, why it was flagged, and gives concrete next steps — change your password, revoke third-party/AI app access, and review recent posts.

Honest limits

Guardian AI is a best-effort, first-level guardrail. It can miss incidents, raise false alarms, or be delayed by third-party platform limits. It is not insurance and not a substitute for two-factor authentication or good security hygiene. See our Limitation of Liability for the full picture.