Legal
AI Notice
How Liralang uses generative AI.
Last updated: 2026-06-26
Where AI appears
Liralang uses AI for conversation, writing correction, grammar search, personalized review, tutoring, and dynamic audio. Some features also use rules and content written by Liralang.
What is sent to providers
To respond, we may send providers such as Anthropic and Inworld the text you write, lesson context, level, base language, conversation memory, vocabulary, or relevant preferences. These providers process what is sent to deliver the service, under their own terms and any data processing agreements we have with them. We log usage telemetry such as model, tokens, duration, estimated cost, and feature.
Voice audio you send for transcription (dictation or speaking) is processed by an external voice provider and may be considered biometric data in some jurisdictions. Using voice is optional: you can use Liralang by typing.
Limitations
- AI can make mistakes, invent explanations, or miss nuance.
- We do not make decisions with legal or equivalent effects solely by AI.
- Corrections are learning help, not official level certification.
- Do not include secrets, documents, health data, payment data, or sensitive third-party information.
User control
You can edit or delete learning content in the app, export your data, and delete your account. If you want us to review a problematic response or data processing issue, contact privacy@liralang.com.