True Member Verification β€” Camera-Based Face Matching, Not Just a Badge Toggle

The problem

Right now, "verification" in BuddyBoss (via Verified Member add-ons or custom profile field hacks) is really just an admin-controlled badge. It tells members "an admin flipped a switch," not "this person is who they claim to be." For communities where trust actually matters β€” marketplaces, dating-adjacent niches, local groups, professional networks β€” a manual toggle doesn't stop catfishing, duplicate accounts, or stolen profile photos. Admins are left eyeballing avatars or asking members to email a photo, which doesn't scale and creates liability around handling those images manually.

The request

A native identity verification flow where a member proves they match their own profile photos using their device camera:

  1. Member opts into verification from their profile (or is required to verify before unlocking certain actions β€” posting in groups, messaging, listing items, etc.).
  2. BuddyBoss prompts a live camera capture (selfie), ideally with basic liveness detection (blink/turn prompt) so a photo-of-a-photo fails.
  3. The capture is compared against the member's profile photo(s) using face-matching. On a match above a confidence threshold, the member is automatically granted verified status; borderline results drop into an admin review queue with the comparison side-by-side.
  4. Verified status displays as the existing badge, but now it actually means something.

Admin controls that should ship with it

Configurable confidence threshold and re-verification interval (e.g., re-verify if the profile photo changes), a manual review queue for failed/borderline matches, the ability to gate features behind verified status (BuddyBoss access controls already have the right hooks for this), and a choice of processing backend β€” either an on-server open-source model or a pluggable third-party API (AWS Rekognition, FaceIO, etc.) for sites that don't want to self-host ML.

Privacy considerations (please build these in, not bolt them on)

Verification captures should be processed and then discarded by default β€” store only the pass/fail result and timestamp, not biometric data. Explicit consent screen before camera access, GDPR/BIPA-friendly defaults, and a member-facing option to delete their verification record.

Why this matters for BuddyBoss specifically

Trust and safety is becoming a differentiator for community platforms. Circle, Mighty Networks, and Discord all lean on external verification bots or nothing at all β€” a native, privacy-respecting face verification flow would be a genuine competitive advantage, especially for the marketplace and membership use cases BuddyBoss already targets. It also reduces admin workload dramatically compared to manual badge management.