Member Insights
Modern security architecture does not guarantee secure implementation. Telecoms operators need to test their controls against realistic attacks, and keep testing as the network changes. SecurityGen Co-founder Dmitry Kurbatov explains.

Telecoms security needs proof, not paperwork
Ask a telecoms operator whether their network is secure, and they will probably point to their latest security assessment. Fair enough. But what happens after the next software upgrade, configuration change or new network function deployment?
An audit can establish whether security controls are in place, but it cannot guarantee they will work tomorrow, or even today, against an attack nobody has tested. The challenge is not simply identifying security gaps. It is knowing whether protection works – and continues working as the network changes.
The telecoms industry has accumulated decades of security knowledge. 3GPP defines mobile network security architecture and requirements. GSMA's Mobile Cybersecurity Knowledge Base, including FS.31, provides security guidance and baseline controls across network domains. MITRE FiGHT describes the tactics and techniques attackers can use against 5G networks.
Together, these resources help operators understand what should be protected, how security should be implemented and what could go wrong. But none of them, on their own, proves that an actual network is secure.
Take 5G Network Repository Function (NRF) discovery. Access to service information should be controlled, yet our assessment teams have encountered implementations where network function information could be retrieved without the expected authentication. The architecture may look right, but the implementation can tell a different story.
Modern infrastructure does not eliminate legacy exposure, either. During a recent webinar, we described an assessment of a newly deployed 4G/5G network whose engineers questioned the need for SS7 testing. Three days later, multiple vulnerabilities had been identified through its roaming interconnections with older networks.
Both examples illustrate the same problem: Modern security architecture does not guarantee secure implementation. Whether it is 5G Service-Based Architecture (SBA) authentication, network function authorization or roaming policies, controls need to be tested against realistic attack scenarios.
Telecoms networks change continuously. Security testing often does not. A new network function, an SBA configuration change, a firewall policy update or a software upgrade can introduce exposures that were not present during the last assessment.
Running another penetration test after every change is rarely practical. But leaving controls untested until the next annual assessment is hardly a solution either. Operators need a way to validate security more frequently, particularly when changes affect critical network functions or previously identified vulnerabilities.
This is where breach and attack simulation (BAS) becomes valuable. BAS does not replace penetration testing. Its purpose is to make relevant attack scenarios repeatable, measurable and available whenever the network changes. Rather than organizing a separate assessment each time, operators can incorporate automated testing into regular network operations.
For example, operators can:
Continuous validation becomes even more valuable when integrated directly into threat detection and remediation. This was one of the approaches demonstrated by Moch Sholeh from Telkomsel through Telco AEGIS, a TM Forum Catalyst project that SecurityGen participated in.
Telco AEGIS connects telecom-specific threat detection, AI-assisted analysis, security policy generation, attack simulation and mitigation into a coordinated workflow. Instead of stopping at identifying suspicious activity or recommending a protective action, the process includes verification of whether the proposed mitigation will actually work.
The project demonstrated this through a subscriber-location tracking scenario. When an intrusion detection system (IDS) detects a suspicious traffic, the system analyzes the threat and checks whether existing firewall controls can block it. If protection is insufficient, a new rule can be proposed and tested using breach and attack simulation (BAS) before deployment. Once the rule is applied, the same attack is simulated again to verify that the firewall now blocks it.
The demonstrated workflow also includes human authorization for security policy changes, an important safeguard when automated decisions affect production networks. This approach allows operators to introduce automation without immediately handing every security decision to an autonomous system.
Deploying a firewall rule proves that an action was taken. Replaying the attack provides evidence that the action worked. The demonstration used common telecom attack techniques aimed at identifying subscribers and tracking their locations, showing how detection, protection and validation can work together in a continuous security process.
A fully autonomous security system is not a prerequisite for continuous validation. Operators can begin with a single high-risk attack path, identify the controls intended to prevent it and establish repeatable tests. These tests can then become part of existing assessment, remediation and change management processes.
There is also a practical consideration: Telecom security requires specialized knowledge. Automation can make testing more frequent and consistent, but teams still need to understand the protocols, vulnerabilities and operational implications of the results. For operators without sufficient internal expertise, an initial period of managed testing and knowledge transfer may be a more realistic starting point than deploying tools without the skills to use them effectively.
Once repeatable validation is established, automated detection, decision-making and response can be introduced gradually. Telco AEGIS demonstrates how these capabilities can work together, while broader adoption will require reliable integration, operational safeguards and further testing across different network environments.
Operators do not need to wait for fully autonomous security to start improving their defenses today.
Security audits remain essential. They provide structure, accountability and a common language for discussing risks. Threat frameworks help operators understand how those risks could be exploited, while Breach and Attack Simulation provides practical evidence of whether the controls intended to stop those attacks actually work.
When the same tests are repeated over time, operators can measure something more meaningful than their latest audit score: whether their defenses are becoming more effective. This also makes it easier to prioritize remediation, verify improvements and identify weaknesses introduced by subsequent network changes.
The goal is not simply to demonstrate compliance. It is to make telecom security measurable, manageable and continuously improving.