“Due to the dynamic nature of cyber-attacks, there are no infallible cybersecurity systems,” said Edgard Capdevielle, chief executive at Nozomi Networks. “However, the risk can be greatly reduced by implementing a layered defense involving anomaly detection with machine learning capabilities where a baseline of industrial control systems can be established and any deviations can be alerted and acted upon. Introducing machine learning and artificial intelligence into the ICS environment is key to faster and more efficient processes for securing unique industrial networks. Finally, closely following the NIST framework and best practices can also improve the risk posture of industrial control systems as standardization helps to facilitate peer-validated security architectures, protocols and guidelines.”

Read More