Common Risks of Machine Learning
28 Oct 2021
In order to cater to the demand for complex interconnected systems, Machine Learning (ML) systems have evolved from traditional IT systems, which are typically simple programmes that handle simple datasets. ML systems create more opportunity by enabling organisations to process more complex use cases which are typically done by humans.
FEATURES OF ML SYSTEMS
MITIGATING THE RISKS OF USING ML SYSTEMS
The table below describes key risk areas of using ML systems and possible mitigation measures to reduce associated residual risks.
KEY CYBERSECURITY FACTORS AND APPROACHES
ML systems process and distil massive amounts of data, reinforcing the need for fundamental cybersecurity measures to protect data confidentiality and integrity, as well as to maintain the availability of ML services. Organisations may consider the following key factors and cybersecurity approaches: