"OpenRouter presents a high-risk privacy profile (71/100) comparable to mainstream ad-supported services like Netflix/Spotify but with additional critical concerns. The service collects extensive data including biometric information and all user inputs to AI models, then shares this data with third-party Model Providers for training purposes beyond OpenRouter's control. Most critically, the company reserves unilateral authority to modify its privacy policy retroactively without notice, with changes automatically applying to existing personal data through a forced-acceptance model."
openrouter Privacy Concerns & Scorecard
Unilateral policy modification authority with retroactive application. Company can modify privacy policy at any time without prior notice and changes apply retroactively to existing personal data. Continued use constitutes forced acceptance.
Extensive collection of sensitive data including biometric information, location data, and AI model inputs/outputs. Biometric data (voiceprints, faceprints, voice models) is processed by third-party models with user responsibility for obtaining consent from data subjects.
User inputs and outputs shared with Model Providers for inference processing, with different providers having different retention and training practices. Model Providers may retain and use data for model training unless user opts out, creating ongoing exposure beyond OpenRouter's control.
Uses cookies, pixels, and IP address tracking for interest-based (targeted) advertising across third-party websites. Third-party ad partners automatically receive IP addresses and deploy their own tracking technologies.
Model Providers have independent and uncontrolled data retention practices. OpenRouter explicitly does not control how LLM providers handle data for model training purposes.
No recommended actions at this time.
Automate your privacy
Connect your accounts to TrueTerms to automate these privacy actions and monitor policy changes.
Data Collection & Tracking
Personal Information Collected
Account creation, authentication, communication, billing, and account management
Processing through third-party AI models for inference; may be used for model training by providers
Service delivery, geographic analytics, fraud detection, targeted advertising
Service optimization, debugging, analytics, fraud detection, and abuse detection
Processing through selected AI models; may be used by Model Providers for training and model improvement
Analytics, service improvement, user experience optimization, abuse detection, security
Payment processing, billing, fraud detection, and financial reconciliation
Customer support, account management, marketing communications, and legal compliance
Cross-Platform Tracking
OpenRouter uses cookies, pixels, and IP address tracking to monitor your behavior across third-party websites and services for delivering targeted advertising and marketing purposes. Third-party advertising partners receive your IP address automatically and deploy their own cross-site tracking technologies.
Tracking Methods:
Cookies
Persistent and session cookies used to track user behavior, preferences, and advertising exposure across OpenRouter and third-party websites
Tracking Pixels
Embedded pixels and web beacons used by OpenRouter and third-party advertisers to monitor user behavior and ad interactions across the web
Device Fingerprinting via IP Address
IP addresses collected and used for cross-site tracking, geolocation, device identification, and delivery of targeted advertising by third-party ad networks
Session Replay
PostHog-based session replay technology that records and replays user interactions, mouse movements, and keyboard activity within the OpenRouter platform
Analytics Tracking
Google Analytics and PostHog tracking to monitor user behavior, engagement metrics, and conversion data within the platform
AI & Data Training
openrouter uses your content to train AI
Below is what's used and how (if at all) you can object.
What Content Is Used
User Inputs (Prompts)
Text, images, audio, and video submitted by users to AI models may be used for model training and improvement by the selected Model Provider
User Outputs (Generated Responses)
AI-generated responses and outputs may be used for model training and improvement by the selected Model Provider
Biometric Data
Voiceprints, faceprints, and voice models extracted from audio/video submissions may be used for model training by the selected Model Provider
How to Object
- 1
Select Model Providers that commit contractually to not using inputs and outputs for model training. Review each Model Provider's privacy terms and data practices independently. Different Model Providers have different policies regarding whether they use data for model training and improvement.
Why We Analyzed openrouter
TrueTerms automatically audits privacy policies and data practices using advanced machine learning to keep you informed and protected. This scorecard is based on the latest available public terms of service and privacy policies as of 2026.