Privacy Policy
Effective date: June 18, 2026 · TROLL2 — Troll-Farm & CIB Radar
Short version: TROLL2 analyses Facebook accounts for likely coordinated-inauthentic-behaviour
using only publicly-visible signals — never private timelines, messages, or login-gated content.
It does not collect your name, email, or account credentials. No ads. No third-party tracking. No selling of data.
1. Who We Are
TROLL2 (Troll-Farm & CIB Radar) is a free decision-support tool for identifying likely Facebook
troll-farm and coordinated-inauthentic-behaviour (CIB) accounts, available at
app.troll2.help and as a browser extension
loaded from source. It is not affiliated with Facebook / Meta. A troll flag is a likelihood estimate
produced from public signals — not proof of wrongdoing.
2. Data We Collect
When an account is analysed, the following data is stored on our servers:
- Public account signals — only information visible to any unauthenticated visitor of the profile (e.g. public handle, public posting/coordination patterns). We do not access private timelines, friends lists, messages, or any login-gated content.
- Analysis result — troll score, troll level, explanation, detected patterns, and coordination-cluster membership
- Anonymous session ID — a random UUID generated in your browser; not linked to your identity
- Timestamp — date and time of the analysis (UTC)
If you submit a report (optional), we additionally store the suspect handle, URL, and any
additional info you provide, linked to the analysed account.
3. Data We Do NOT Collect
- Private Facebook content — timelines, messages, friends lists, or anything behind a login wall
- Your name, email address, or phone number
- Your Facebook / Google account or any login credentials
- Your location or IP address (beyond what our hosting infrastructure logs transiently)
- Device identifiers (IMEI, advertising ID, etc.), photos, or contacts
4. How We Use Your Data
- CIB analysis — public signals are processed (including via an AI API such as Anthropic Claude and/or Google Gemini) to produce a troll-likelihood estimate. AI providers process inputs under their own privacy policies.
- Live troll alerts — when an account scores ≥ 0.6, an anonymised alert (troll level and score only) is broadcast to other connected users in real time. No personal data or account names are included.
- Admin review — submitted reports and coordination clusters may be reviewed by an administrator to curate detection patterns.
- Service improvement — aggregate, anonymous statistics are used to monitor health and improve detection accuracy.
5. Data Retention
Analysis records (public signals, results, session IDs) are automatically deleted after 90 days.
A scheduled cleanup runs daily and permanently removes records older than 90 days.
6. Data Sharing
We do not sell, rent, or share your data with advertisers or data brokers. Data is shared only in these limited cases:
- AI providers — public signals are sent to Anthropic and/or Google for analysis. See their privacy policies for how they handle API inputs.
- Legal obligations — we may disclose data if required by law or court order.
7. Security
Our servers run on AWS with encrypted storage, HTTPS-only access, and restricted administrative
access. Rate limiting and bot protection are in place to prevent abuse.
8. Responsible Use
TROLL2 is intended as decision-support, not surveillance. A troll flag is a likelihood estimate
based on public signals and must not be used to harass, dox, or take action against any individual.
Always apply your own judgement.
9. Your Rights
Because we do not collect personally identifiable information, most data subject rights
(access, correction, deletion) are satisfied by the 90-day automatic purge. If you submitted
a report with a specific suspect handle or URL and wish it removed, contact us below.
10. Changes to This Policy
We may update this policy as the app evolves. The effective date at the top of this page
will reflect the latest revision. Continued use of the app after changes constitutes acceptance.
11. Contact