News report: $20M was stolen from Mexican banks, with the initial intention to steal $150M. Automatically we are drawn to think of a “Casa de Papel” style heist, bank robbers wearing masks hijacking a bank and stealing money from an underground vault. This time, the bank robbers were hackers, the vault is the SPEI application and well, no mask was needed. The hackers were able to figuratively “walk right in” and take the money. Nothing was stopping them from entering the back door and moving laterally until they reached the SPEI application.
Central bank Banco de México, also known as Banxico, has published an official report detailing the attack, the techniques used by the attackers and how they were able to compromise several banks in Mexico to steal $20M. The report clearly emphasizes how easy it was for the attackers to reach their goal, due to insecure network architecture and lack of controls.
The bank heist was directed at the Mexican financial system called SPEI, Mexico’s domestic money transfer platform, managed by Banxico. Once the attackers found their initial entrance into the network, they started moving laterally to find the “crown jewels”, the SPEI application servers. The report states that the lack of network segmentation enabled the intruders to use that initial access to go deeper in the network with little to no interference and reach the SPEI transaction servers easily. Moreover, the SPEI app itself and its different components had bugs and lacked adequate validation checks of communication between the application servers. This meant that within the application the attackers could create an infrastructure of control that eventually enabled them to create bogus transactions and extract the money they were after.
Questions arise: what can be learned from this heist? How do we prevent the next one? Attackers will always find their way in to the network, so how do you prevent them from getting the gold?
Follow Advice to Remain Compliant
When it comes to protecting valuable customer information and achieving regulatory compliance, organizations such as PCI-DSS and SWIFT recommend the following basic steps: system integrity monitoring, vulnerability management, and segmentation and application control. For financial information, PCI-DSS regulations enforce file integrity monitoring on your Cardholder Data Environment itself, to examine the way that files change, establish the origin of such changes, and determine if they are suspicious in nature. SWIFT regulations require customers to “Restrict internet access and protect critical systems from the general IT environment” as well as encourage companies to implement internal segmentation within each secure zone to further reduce the attack surface.
Let’s look at a few guidelines, as detailed by SWIFT while incorporating our general advice on remaining compliant in a hybrid environment.
- Inbound and outbound connectivity for the secure zone is fully limited.
Transport layer stateful firewalls are used to create logical separation at the boundary of the secure zone.
- No “allow any” firewall rules are implemented, and all network flows are explicitly authorized.
Operators connect from dedicated operator PCs located within the secure zone (that is, PCs located within the secure zone, and used only for secure zone purposes).
- SWIFT systems within the secure zone restrict administrative access to only expected ports, protocols, and originating IPs.
- Internal segmentation is implemented between components in the secure zone to further reduce the risk.
SPEI servers, that serve a similar function to SWIFT application servers should adhere to similar regulatory requirements, and as elaborated on by Banxico in the official analysis report, such regulations are forming for this critical application.
Don’t Rely on Traditional Security Controls
The protocols detailed above are recommended by security experts and compliance regulations worldwide, so it’s safe to assume the Mexican bank teams were aware of the benefits of such controls. Many of them have even been open about their attempts to implement these kinds of controls with traditionally available tools such as VLANS and endpoint FWs. This has proven to be a long, costly and tiresome process, sometimes requiring 9 months of work to segment a single SWIFT application! Would you take 9 months to install a metal gate around your vault and between your vault compartments? I didn’t think so…
Guardicore Centra is set on resolving this challenge. Moving away from traditional segmentation methods to use microsegmentation that provides foundational actionable data center visibility, this technology shows quick time to value, with controls down to the process level. Our customers, including Santander Brasil and BancoDelBajio in Mexico, benefit from early wins like protecting critical assets or achieving regulatory compliance, avoiding the trap of “all or nothing segmentation” that can happen when competitors do not implement a phased approach.
Guardicore provides the whole package to secure the data center, including real-time and historical visibility down to the process level, segmentation and microsegmentation supporting various segmentation use cases, and breach detection and response, to thoroughly strengthen our client’s security posture overall.
Micro-segmentation is more achievable than ever before. Let’s upgrade your company’s security practices to prevent attackers from gaining access to sensitive information and crown jewels in your hybrid data center. Request a demo now or read more about smart segmentation.