Security by Design: Building Resilience Into Telecom Networks for an AI-Driven Threat Landscape

Artificial intelligence (AI) is rapidly changing the cybersecurity equation. The same technology helping organisations automate operations and improve services is opening up new ways for attackers to scale reconnaissance, generate malicious content, and coordinate sophisticated attacks. This is why resilience is crucial.

For telecommunication providers, resilience matters because networks underpin financial services, government operations, cloud computing, healthcare, transport and everyday digital life. A successful attack creates consequences far beyond the network itself.

According to INTERPOL’s 2025/2026 Asia and South Pacific Cyberthreat Assessment, more than half of the countries surveyed reported that cybercrime accounted for 30% of all recorded crime nationally. The report found that distributed denial-of-service (DDoS) attacks surged 92% in 2024, while discussions about deepfakes on cybercriminal forums and Telegram channels popular among Southeast Asian threat actors increased by 600% between February and June 2024.

The threat landscape is also expanding. AI-powered DDoS attacks, multi-vector tactics, and AI-generated deception put pressure on both technical defenses and the people operating them. An analysis of proactive AI defense for telecommunications highlights how attackers are using AI to automate multi-layered attacks, manipulate communications, and develop malicious code capable of bypassing conventional defenses.

At the same time, the rapid adoption of generative AI is creating new security challenges of its own. Gartner forecasts that 25% of enterprise generative AI applications will experience at least five minor security incidents per year by 2028, up from 9% in 2025.

The implication is clear: relying on security teams to react to alerts after an incident is no longer enough.

A more proactive model is needed, using AI to articulate knowledge, develop detection and response content, identify patterns, and predict threats before they escalate. The goal is to augment, not replace, human judgment, while recognizing that telecommunication providers need security tailored to specialised network architectures, protocols and operational environments.

Resilience Beyond the Perimeter

Telecommunication networks are becoming increasingly distributed, with cloud-native architectures, virtualization, Open RAN (Radio Access Networks), edge computing and software-defined infrastructure creating new connections across network functions, applications, APIs and operational systems. This flexibility also expands the attack surface.

Modern RANs are increasingly software-defined and interconnected. Recent research on five security threats in RAN highlights how this complexity can create vulnerabilities that perimeter defenses alone cannot address. Security teams need to look beyond traditional indicators of compromise, monitoring for unauthorised remote access, lateral movement, privilege escalation, abuse of radio-node access and suspicious binaries within network functions.

At the same time, attackers increasingly target availability, not just data, through DDoS attacks, supply-chain compromises and attacks on shared infrastructure. That makes resilience as important as prevention: networks must be designed to anticipate threats, limit their impact, maintain essential services and recover quickly when disruption occurs.

Quantum Risks

Resilience also means preparing for a threat that may not arrive tomorrow: quantum computing.

Quantum computers could undermine widely used cryptographic standards. Boston Consulting Group’s analysis of quantum computing and cybersecurity estimates that around 2035, quantum computers could become powerful enough to compromise current standards, making the transition to post-quantum cryptography a priority today. The risk is already present: attackers can collect encrypted data now with the intention of decrypting it later, a threat known as “harvest now, decrypt later.”

For telecom providers, AI and cloud providers, and other mission-critical enterprises, preparing means knowing where cryptography is used, identifying systems that will be difficult to upgrade, and building crypto-agility into the network. A quantum-safe networking approach combines crypto-agility, crypto-resiliency and multilayer encryption to help organisations adapt as standards and threats evolve.

The shift is already moving from theory into practice in Southeast Asia. In September 2025, Maxis announced Malaysia’s first QSN solution for government agencies and businesses, supported by Nokia’s QSN capabilities.

The message is broader than quantum: cybersecurity can no longer be measured only by what organisations prevent, but by how well they adapt, withstand disruption and recover.

Resilience Is a Business Capability

For a telecommunications provider, network availability is directly connected to customer experience, revenue, regulatory obligations and national digital infrastructure.

As AI makes attacks faster and more adaptive, organizations need continuous security capabilities that combine proactive threat discovery, intelligent automation, secure cloud-native architectures and resilient network design.

We should therefore ask not simply, “How do we prevent the next attack?” but, “If an attack succeeds, how quickly can we detect it, contain it, maintain essential services and recover?”

That mindset will become increasingly important as networks evolve toward AI-native 6G. Security cannot be bolted on after the architecture is built. It must be embedded throughout the network cycle, with resilience, reliability, privacy and automated security treated as fundamental requirements. Security research for AI-native 6G highlights the importance of a secure-by-design approach incorporating resilience, reliability, automated security and quantum-safe cryptography.

Nokia’s participation in the NVIDIA Open Secure AI Alliance is one example of how technology leaders are helping the industry advance secure AI infrastructure and build trust into AI-enabled networks.

In the AI era, the strongest network will be the one that can see the threat coming, absorb the disruption, recover quickly, and keep the services that society depends on running.