← back to ideas

AnthropicGuard

7.8
security profitable added: Saturday June 2026 17:52

A security auditing software that analyzes large language model interactions to detect potential adversarial prompting and unauthorized data extraction, inspired by concerns around Alibaba's model querying of Claude.

180h
mvp estimate
7.8
viability grade
10
views

technology stack

Python Medium PostgreSQL

inspired by

Anthropic accuses Alibaba of copying Claude.