AI Anti-Poison Shield
7.8
A software platform designed to detect and filter 'poisoned data' aimed at corrupting AI models, specifically addressing the threat highlighted by underground resistance groups. It provides a real-time data validation layer for AI training pipelines, flagging and mitigating anomalous data points before they impact model accuracy.
180h
mvp estimate
7.8
viability grade
31
views
technology stack
Python
Difficult
PostgreSQL
inspired by
Underground Resistance Aims To Sabotage AI With Poisoned Data