ITU standard outlines how machine learning can strengthen data security in big data systems
A new ITU recommendation explains how machine learning can be used to detect threats and protect data across large-scale digital infrastructures.
The International Telecommunication Union (ITU) has published a new technical standard, ITU-T X.1753, which provides guidance on how machine learning can be used to improve data security in large-scale data systems.
To understand the issue, it is important to consider how modern digital systems work. Many organisations rely on ‘big data infrastructure’, meaning systems that collect, store, and process large volumes of data from different sources. These systems are used across cloud services, public administration, and online platforms.
Because of their size and complexity, these systems face multiple security risks. Data can be exposed during collection, storage, processing, or management. Threats may come from external attackers, system vulnerabilities, or even internal misuse.
The ITU recommendation explains that traditional security methods, such as access controls or encryption, are not always sufficient on their own. This is where machine learning is introduced as an additional tool.
Machine learning refers to systems that can analyse large amounts of data and identify patterns. In the context of security, it can be used to detect unusual behaviour, predict potential threats, and respond more quickly to risks.
The document outlines several practical uses. These include identifying and classifying data, detecting unauthorised access, monitoring system activity, and improving data protection over time. For example, machine learning can help identify sensitive data, detect abnormal user behaviour, or adjust security settings based on changing risks.
It also describes how machine learning can support different stages of data protection. These range from authentication and access control to auditing, backup, and secure deletion of data.
In addition, the recommendation sets out general principles for using machine learning in this context. These include ensuring compliance with laws, maintaining data reliability, avoiding biassed or unfair outcomes, and enabling systems to explain their decisions.
The standard is intended as guidance rather than a mandatory requirement, but it provides a framework for organisations looking to integrate machine learning into their data security practices.
