Principles for the security of machine learning

Overview

These principles aim to be wide reaching and applicable to anyone developing, deploying or operating a system with a machine learning (ML) component. They are not a comprehensive assurance framework to grade a system or workflow, and do not provide a checklist. Instead, they provide context and structure to help scientists, engineers, decision makers and risk owners make educated decisions about system design and development processes, helping to assess the specific threats to a system.

This content is available under the Open Government Licence v3.0

External Links