When 5G powers millions of devices at home, in industries and across our communities, decisions will need to be made in fractions of a second as to where to send a device’s traffic to maintain uptime and provide a superior user experience. In addition, as 5G services are delivered from the broader communications ecosystem, the complexity of real-time data streams from different sources increases exponentially.

As CSPs increasingly onboard enterprise solutions, they will need greater intelligence into what is happening in their networks. AI, machine learning and automation will be required to process and manage the volume and complexity of data arising with 5G networks. These new capabilities will also improve the ability for CSPs to maintain and troubleshoot network performance and to detect network security threats and fraud instances— including new and unknown ones. In fact, Gartner estimates that by utilizing machine learning within their fraud management systems, CSPs can reduce fraud losses by 10 percent due to the technology’s ability to uncover ‘unknown, unknowns’ and the speed in which ML can help detect automated attacks.

Where AI and machine learning help CSPs understand ‘what’ is happening in the network, Data-Science-as-a-Service will go one step further by helping CSPs determine the ‘why.’ This will bring critical intelligence to the network by contextualizing the network’s burgeoning data lake of on-network and partner data to help optimize it, provide a 360^○ view of the customer and get a complete view of the underlying threats.

Network Intelligence for Real-Time Risk Management

The infinite number of potential use cases for connecting things to the Internet and the growing ecosystem of players will bring added complexity that will increase the opportunity for bad actors to find new ways of breaching network security and committing fraud. For example, IoT and eSIMs create new opportunities for traditional types of telecom fraud—such as subscription fraud, international revenue share fraud (IRSF) and traffic pumping—to make a resurgence. Add to this the growing use of digital channels and CSPs are left vulnerable to new types of online fraud such as synthetic identity fraud and account takeover.

With 5G, it becomes more apparent that CSPs will require robust and comprehensive risk mitigation methodologies and auditing capabilities for billing and rating validation, fraud detection and prevention and network security. This is only possible with the power of an integrated end-to-end risk management strategy that spans fraud management, revenue assurance, business assurance and network security. When equipped with advanced analytics and automation, CSPs have the intelligence to turn data into powerful risk management insights and actions in order to predict, detect and safeguard against vulnerabilities in real time. In the 2020s, policy-driven rules built on top of OSS/BSS data will have to work side-by-side with more complex analytics solutions and advanced anomaly detection systems to drive real value and protect the ecosystem.

A recurring opportunity and challenge for CSPs in the 2020s will be how to leverage their data in the network to deliver intelligent 5G monetization, to build intelligent security mechanisms and to deploy intelligent real time risk management strategies. Capabilities such as AI, machine learning and data science will be critical to manage the volume, variety and velocity of data in a 5G world and will be critical to ensure that CSPs extract the most value out of their 5G investments—and to make sure they are able to safeguard their customers from growing security and fraud threats.