# PaperTLDR Privacy Policy

Last updated: 2026-04-19

PaperTLDR collects the minimum data needed to send a daily academic digest or record paid-plan interest: email address, subscribed arXiv categories, optional keywords, active subscription tier or waitlist tier, generated digest records, and operational usage data such as LLM cost events.

PaperTLDR does not need sensitive personal information. Do not submit private research data, health information, student records, passwords, API keys, or confidential manuscripts as keywords.

Email delivery is designed for Resend. Paper classification and summarization are designed for Claude Haiku 4.5 through Anthropic. In this MVP, tests and demo scripts use mocks or dry-run clients. A production operator must review vendor data processing terms before enabling live API keys.

Data can be deleted by removing the subscriber or pricing-interest record from the SQLite database. Digest and usage retention should be set by the production operator before launch.