2
That was a central message from panellists at Industrial Wireless Forum last week – on a session called, “Cybersecurity by design – from secure devices to quantum-safe networks”.
In sum – what to know
Security starts at design – Panelists said cybersecurity must be addressed during system development, with risk analysis and security decisions incorporated before designs are approved.
5G raises new challenges – Connecting industrial equipment and automated guided vehicles to 5G networks introduces additional security considerations, particularly when connectivity supports control functions with safety implications.
AI cuts both ways – Threat actors are using AI to accelerate their understanding of industrial systems, while defenders can use it to identify vulnerabilities, recognize attack patterns and strengthen their responses.
Industrial companies need to treat cybersecurity as a fundamental design requirement rather than an additional layer applied after deployment, particularly as operational technology (OT) environments become more connected and complex. That was a central message from panelists at the RCR Wireless News Industrial Wireless Forum 2026 session, “Cybersecurity by design: From secure devices to quantum-safe networks.”
The discussion brought together Scott Cadzow, chair of ETSI TC SAI (Securing Artificial Intelligence) and standards expert at ETSI; Peter Davies, technical director at Thales e-Security; and Sameer Koranne, head of OT Security Cyber and Data Resilience at Kroll.
Koranne said industrial cybersecurity must prioritize operational safety and continuity because industrial control systems often support essential services. He identified three fundamental areas for organizations to address: strong authentication, secure communications, and secure configuration.
“Identity is the new perimeter,” Koranne said, arguing that organizations need to move away from shared access and shared credentials. He also highlighted the importance of securing communications between systems and ensuring that vendors help maintain secure configurations over the operational life of their equipment.
Davies emphasized that industrial systems evolve over time, meaning their design parameters and boundaries can change. Organizations therefore need to maintain justifiable confidence that their systems will function as expected throughout their life cycles.
He also pointed to the growing complexity of integrated systems and supply chains, particularly in environments involving technologies such as 5G. Understanding the security of suppliers, as well as that of the systems and components they provide, must therefore form part of the overall cybersecurity-by-design approach.
Cadzow argued that security is frequently considered too late in the development process. “You can’t add effective cybersecurity,” he said. “We need to push the idea that you cannot have a design signed off unless you’ve addressed cybersecurity.”
Cadzow said companies should conduct risk analysis, carry out the necessary work and justify their security decisions before completing a design. He also pointed to legislation, including the Cybersecurity Act and Cyber Resilience Act, saying that these measures are pushing companies to address security if they want their products to reach the market.
5G brings security closer to industrial operations
The panel also examined the challenges of introducing 5G connectivity into industrial environments, where networks may support applications that directly influence physical operations.
Davies said organizations need to understand the outcomes they expect from a network and establish whether they can achieve them reliably. In industrial settings, this includes considering how network resources are delivered, rather than simply whether connectivity is available.
For safety-critical applications, timing and service prioritization can become important security considerations. Davies gave the example of equipment controlling when an industrial press operates, explaining that such functions must work within specific time constraints and take precedence over other network traffic.
He also highlighted the difficulty of validating advanced 5G technologies, including machine-learning algorithms involved in beamforming. In addition, he noted that some parts of an industrial system may not map into the communications or industrial security operations centers used to monitor it.
Koranne said the security implications depend partly on how much access organizations provide through 5G networks and which decisions they allow those networks to support. One relatively straightforward application involves remote workers in a plant collecting equipment reliability information through a connected 5G network.
The challenges become more significant when 5G is used to control or guide automated guided vehicles (AGVs). In that situation, connectivity is no longer used only to collect operational data; it can influence the actions of physical equipment and create potential safety implications.
Koranne said companies need to consider how devices will be protected, how communications will take place, how devices will be registered and authenticated, and how data and control commands will be secured.
He emphasized that organizations need to consider the security of communications protocols and authenticate devices, not just individual users, when industrial networks support control applications.
Cadzow added that designers need to consider how systems could be attacked, rather than focusing exclusively on their intended functionality. He argued that security should receive greater attention during software and hardware development, alongside functionality and performance.
AI creates new threats – and defensive opportunities
The panel also discussed how artificial intelligence is changing the cybersecurity threat landscape and how industrial companies could use the technology to strengthen their defenses.
Koranne said there is documented evidence of threat actors using AI to improve their capabilities, including researching industrial systems and developing malware. He argued that AI is reducing the time attackers need to understand how industrial control systems operate, weakening what had previously been a barrier to targeting industrial environments.
He called for AI-related risks to be incorporated into industrial cyber process hazard analysis, alongside the safety checks conducted when industrial control systems are designed. This includes examining how AI systems are authenticated and how their decisions could affect safety.
Cadzow, meanwhile, said AI can help defenders analyze data quickly, identify patterns associated with attacks and find weaknesses in their own systems. He argued that organizations should use these capabilities to identify and address vulnerabilities before attackers exploit them.
He also called for more widespread use of red teams and blue teams to test systems and improve defenses. However, he stressed that this work must be continuous because attackers also adapt and learn.
Click here to access the sessions of the Industrial Wireless Forum 2026.
——————————————————-
Click Here For The Original Source.
