Live meeting deepfake protection for Zoom, Teams and Meet

Impersonation on video calls, from fake executives authorising payments to fake candidates in interviews, is one of the fastest-growing deepfake risks. sealverity.ai lets an analyst paste a meeting link and send a bot that checks every participant while the call is happening.

How it works

An analyst starts a live check from the app and pastes a Zoom, Microsoft Teams or Google Meet link. The sealverity.ai bot asks to join; once admitted, it posts a consent notice in the meeting chat. Where an organisation has made the notice mandatory, it is enforced on the server for every bot, however the session was started.

Audio and video are streamed in chunks of about 10 seconds per participant and scored by the detector ensemble. The live session page shows a risk gauge for each participant, a timeline and per-detector scores for each chunk.

Works with Zoom, Microsoft Teams and Google Meet
Mandatory consent notice in the meeting chat
Per-participant gauges updated every 10 seconds
Expandable per-detector scores for every chunk
Separate live-call threshold and detector weights

Alerts are review items, not verdicts

An alert is raised only after two consecutive chunks for the same participant exceed the organisation's live-call threshold. Alerts are worded as possible synthetic media that needs analyst review and are placed in the review queue. Notifications can go in-app, by email or to a webhook.

Live calls use their own threshold, set per organisation with a recorded reason, and their own detector weights, so teams can reduce the influence of detectors that over-react to compressed meeting audio.

After the call

When the check ends, the bot leaves the meeting and a post-call evidence record is created with participants, chunk scores and alerts. It can be attached to a case and included in a signed forensic report alongside other evidence.

Controls and privacy

Live checks are started only by people in your organisation; bots are never sent automatically. Sovereign mode is respected, so organisations that block external detectors are not exposed through live calls. All actions are recorded in the audit log, and usage is billed per minute analysed.

Typical uses

Finance teams verifying who is really on a payment-approval call, HR teams running remote interviews, executive protection for board and investor meetings, and investigators who need a defensible record of a suspicious call.

Part of one investigation platform

This capability is part of sealverity.ai, a single platform for media authenticity and investigations. Findings from images, video, audio, live calls, identity checks and social and news monitoring can be linked to cases, entities and a link graph, with risk scores that show which factors contributed. Access is protected by mandatory multi-factor authentication, single sign-on and SCIM provisioning, five roles, retention policies and legal hold. Organisations can bring their own self-hosted detectors or switch on sovereign mode, which blocks every external call on the server and logs each blocked attempt. Integrations include a browser extension, Slack and Teams, email scanning, webhooks, SIEM connectors and a REST API.