Migrating banks’ old technology to cloud computing systems is creating a cyber security nightmare for their IT and risk teams, experts have warned.

Cloud infrastructure — which enables software and data to be held off-site and accessed by any part of an organisation, in any location — is helping the world’s banks develop digital services with greater ease and speed. 

But experts warn that cloud adoption can also be highly risky for banks and financial groups, because cyber criminals are increasingly exploiting security “holes” and misconfigured settings in cloud platforms to steal data, defraud customers and disrupt operations. 

With so much sensitive financial and personal information stored in the cloud, as digital banking has become widely adopted, data breaches have become a leading concern for financial groups, says Simon Crocker, senior director of systems engineering for western Europe at cyber security company Palo Alto Networks.

Cyber criminals’ approach is to access data by identifying vulnerabilities and misconfigurations in the cloud services used by banks, he explains. But the damage does not stop there; criminals can also take over customer accounts, commit financial fraud and access other banking resources, he adds.

“One of the key threats banks face when securing their cloud environment is attackers gaining unauthorised access through their inbound traffic, such as a customer’s online banking transactions or account opening, or outbound traffic, which includes activities such as payment processing, trading and interbank communication,” Crocker explains.

He says hackers can intercept banking traffic by launching distributed denial-of-service (DDoS) attacks, which overwhelm computer servers with large volumes of requests, as well as SQL (structured query language) injection and cross-site scripting attacks, which inject malicious code into applications and websites.

Moving to cloud systems is, therefore, a trade-off in terms of risk, Crocker argues. “Ultimately, banks rely on cloud service providers to deliver secure and reliable infrastructure and services,” he says. “However, vulnerabilities in cloud platforms, misconfigurations, or security weaknesses in the underlying infrastructure can expose banks to significant security risks.”

Banks’ best options for mitigating these is using encrypted communication methods, such as virtual private networks, dedicated private connections and web proxy servers, Crocker advises. In addition, he recommends using network segmentation — whereby computer networks are separated into smaller parts — to limit the impact of security breaches and improve control over outbound traffic flows.

A bigger challenge, however, is dealing with unknown, or unaddressed, security flaws that bank IT teams have not prepared for — known as “zero-day” exploits.

“‘Zero-day’ describes recently discovered security vulnerabilities that hackers can use to attack systems, and literally means that an organisation under attack has ‘zero days’ to fix it,” explains Sergey Lozhkin, principal security researcher at Russian antivirus software company Kaspersky, and former vice-president of cyber security operations for JPMorgan Chase. He warns that these can give cyber criminals a strong foothold in cloud banking systems.

Advanced persistent threats, or APTs, are attacks that can also go undetected, giving hackers a big advantage. “APTs . . . exploit vulnerabilities to gain prolonged access to a bank’s cloud infrastructure, allowing them to exfiltrate sensitive data over time,” says Lozhkin.

While both types of attack present significant cyber security risks, the sophisticated nature of APTs is forcing banks to step up their defences. Lozhkin points out that APT hacking techniques were instrumental in the 2016 cyber attack on South Africa’s Standard Bank Group, when cybercriminals stole $13mn by forging 1,600 cards. “APTs are like a stealthy burglar that can sit within networks completely unseen for any length of time before striking,” he says. 

To mitigate zero-day exploits, Lozhkin recommends the use of advanced monitoring solutions to detect “unusual activities indicative of a zero-day attack”. Automated tools can also “streamline this process” and decrease “the window of opportunity for attackers”, he adds. 

And, to mitigate APTs, he suggests threat detection solutions, network traffic analysis, and endpoint detection and response (EDR) systems, which provide continuous cyber security monitoring and user behaviour analytics (UBA).

Lozkhin is confident this approach will work. “By implementing a comprehensive security strategy that includes regular updates, configuration management, advanced threat detection and robust incident response plans, banks can mitigate the risks posed by zero-day exploits and advanced persistent threats, protecting their assets and maintaining customer trust in an increasingly digital world,” he says.

However, Jake Moore, UK-based global cyber security adviser at Bratislava-headquartered security company ESET, fears many institutions are not taking action quickly enough. “The banking industry has been slow to adopt cloud security and this has been made slower due to the tough regulations the industry faces,” he says. “Ransomware attacks, which now commonly include data-compromising techniques, pose one of the most significant risks to financial institutions.”

According to Moore, implementing a “multi-layered security approach” will help banks mitigate these risks. These layers should include stringent authentication protocols — such as physical security keys, unique passwords and device identifiers — to prevent unauthorised persons from accessing cloud systems. 

Regular security audits will also help IT teams find and fix vulnerabilities in cloud-based banking systems, he suggests, while strong encryption can make sensitive data unreadable, even if it is stolen by cyber criminals.

But, with human error being the cause of most cloud security breaches, according to a report by defence group Thales, Moore urges banks to train their staff in tackling cyber threats.

Many can be mitigated using zero-trust models, says Tristan Morgan, managing director of security at telecoms group BT. These demand that everyone trying to use a bank’s WiFi network, whether employee or customer, is “constantly checked and validated”.

“It also provides visibility of who is on the network, reducing risks, and supporting the operational needs of companies in a hybrid working environment,” says Morgan. 

Bernie Wright, chief information security officer at cloud infrastructure provider ClearBank, advises banks to operate a “comprehensive supplier onboarding process”, to eradicate security risks in their supply chains. He notes that many suppliers offer software-as-a-service products (licensed on a subscription basis) that are run in the cloud and, if improperly secured, can provide hackers with backdoor access to banking clients’ IT environments. 

“There are certain levels of trust that are needed so, as part of due diligence, it’s crucial to review how suppliers operate, their associated corporate policies, and resilience capabilities,” Wright emphasises. 

However, far greater cloud computing applications — and threats — are expected to arise in coming years from quantum computers. These devices harness quantum mechanics to carry out vastly more, and faster, processing operations than today’s computers could ever manage.

Kamran Ikram, senior managing director in financial services at the UK and Ireland arm of consultants Accenture, sees both pros and cons. “Banks can build more resilient and secure financial systems with quantum algorithms constructed to find opportunities for credit scoring and optimising trading trajectories,” he says. “But quantum computing will also allow encryption codes to be cracked in a fraction of the time they now take.” 

Copyright The Financial Times Limited 2024. All rights reserved.
Reuse this content (opens in new window) CommentsJump to comments section

Follow the topics in this article