Plain language summary: Candor is an open social feed with AI-powered safety systems running in the background. We don't sell your data. We don't censor opinions. Our safety tools run silently to keep the feed honest. By joining, you agree to let us run these systems.
Candor: The Open Feed Network is operated by Open Feed Network, Inc., a Wyoming C-Corporation (EIN: 42-2688100). By creating an account, you enter into a binding agreement with Open Feed Network, Inc.
By creating an account, you confirm that you have read, understood, and agree to these Terms and our Privacy Policy. You must be at least 13 years of age. We record your consent at signup including IP address, browser info, version accepted, and timestamp.
Candor operates AI-powered safety modules extracted from the Candor Reach Gateway — the same systems protecting the Candor feed, made available as commercial services. By using Candor, you explicitly consent to your content being processed by these systems. These include:
Standalone open-source disinformation detection engine. Analyzes post content and assigns transparency scores in real time. Scores are informational — Truth Shield does not automatically remove content. Free, open source, and available for any platform. Bilingual English and Spanish.
Advanced journalism edition of Truth Shield combined with secure tip submission infrastructure. Provides anonymous, encrypted tip intake with dual-engine analysis: Claude Haiku for fast batch scoring and Claude Sonnet for deep analysis with citations and influence campaign detection. Built for newsrooms, independent journalists, and investigative reporters.
Live mental health distress and suicide prevention software. Monitors the feed in real time for signals of mental health distress and escalates through a structured response ladder. A trained human dispatcher makes all critical response decisions. No automated actions are taken at critical intervention stages.
Live mental health distress and suicide prevention software built specifically for first responders — police, fire, EMS, and dispatch personnel. Uses discipline-specific detection models. Alerts routed to peer support officials, never to HR or chain of command. Architecturally designed to protect employment.
Counter-terrorism and violent extremist content (TVEC) detection engine. 7-stage detection pipeline analyzing all content in real time. Confidence score of 85 or above triggers immediate quarantine, law enforcement preservation, and automatic reporting to the FBI Internet Crime Complaint Center (IC3). Novel TVEC hashes contributed to GIFCT upon membership approval. Bilingual English and Spanish. Live stream detection window of 4-8 seconds.
Bot detection and active defense system. Operates silently on every request. Detects automated behavior using behavioral fingerprinting, request signature analysis, and timing patterns. Issues proof-of-work challenges to suspicious traffic. Human users are not affected.
7-layer child internet safety gateway for the adult Candor platform. Detects minors attempting to access adult content using linguistic analysis, behavioral patterns, profile signals, network graph analysis, proof-of-humanity, zero-knowledge age verification, and community reporting. Response is always protective, never punitive. Compliant with COPPA, GDPR-K, and KOSA.
Inverted child safety gateway operating on the Your Voice youth platform (yourvoicesnetwork.org). Uses the same 7-layer detection system in reverse — identifying adults and potential predators attempting to access the youth platform. Detects grooming language patterns, adult behavioral signals, and systematic targeting of younger users.
Real-time broadcast integrity monitoring for live streams and podcasts. Converts audio to text and analyzes every claim as it is spoken — returning verdicts of verified, unverified, disputed, opinion, or satire with confidence scores and source citations. Detects paid influence campaigns and undisclosed financial relationships. Bilingual English and Spanish, verified in production.
The master intelligent safety routing layer controlling all requests entering the Candor platform. Manages authentication, rate limiting, attack defense, and routing across all Candor services. Sentinel bot detection is embedded within the Gateway. Available as a managed service for other platforms that need intelligent safety feed routing.
All safety systems are operated solely to maintain feed integrity and user safety. Not used for advertising, data monetization, or political targeting.
You retain ownership of all content you post. By posting, you grant Open Feed Network, Inc. a non-exclusive, royalty-free, worldwide license to store, display, and distribute your content through the platform.
The following content is prohibited:
Candor does not prohibit opinions, political speech, satire, or controversial viewpoints.
Candor explicitly prohibits terrorist and violent extremist content in all forms, including:
This prohibition applies to text, images, video, audio, and live streams in all languages including English and Spanish.
Candor's ThreatShield system analyzes all content in real time against TVEC signals using a 7-stage detection pipeline. Confirmed TVEC is immediately quarantined, preserved for law enforcement with a full chain of custody, and reported to appropriate authorities including the FBI Internet Crime Complaint Center (IC3). Novel TVEC hashes are contributed to industry hash-sharing databases upon membership approval.
Candor is committed to the standards of the Global Internet Forum to Counter Terrorism (GIFCT) and is actively pursuing GIFCT membership to participate in the shared hash database and Crisis Information Protocol (CIP). We comply with GIFCT transparency, human rights, and civil society engagement criteria and publish annual transparency reports at candortheopenfeednetwork.com/transparency.html.
Journalism, academic research, counter-extremism work, and survivor testimony that references TVEC without promoting it are explicitly protected. Our detection systems are designed to distinguish TVEC promotion from these protected categories.
We do not sell your data. We do not use your data for advertising. Full details in our Privacy Policy.
The core feed algorithm and Truth Shield are open source at github.com/OpenFeedNetwork/platform. Sentinel, Gateway, Shield, and Beacon are closed source to prevent circumvention.
These terms are governed by the laws of the State of Wyoming, United States of America.
Legal: [email protected]
Open Feed Network, Inc. · Wyoming C-Corp · EIN 42-2688100