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
5
views
technology stack
Python
Difficult
PostgreSQL
inspired by
Underground Resistance Aims To Sabotage AI With Poisoned Data