Skip to main content Skip to footer
  • "com.cts.aem.core.models.NavigationItem@67263327" Careers
  • "com.cts.aem.core.models.NavigationItem@7df8c8df" News
  • "com.cts.aem.core.models.NavigationItem@5e3465a0" Events
  • "com.cts.aem.core.models.NavigationItem@6ee4adc1" Investors


June 18, 2025

Cybersecurity in banking, 2025: The critical gaps to address now

Six urgent shifts banks must make now to avoid catastrophic cybersecurity breaches in 2025 and beyond.

Visit our Banking webpage

Cybercrime has become a global crisis, with annual costs spiraling to a predicted $10.5 trillion this year. Financial institutions are prime targets for these highly sophisticated AI-driven cyberattacks, ransomware disruptions and looming quantum-powered encryption breaches. Yet, many banks are still operating security models designed for yesterday’s threats. Banks need a radical cybersecurity transformation—not just incremental updates.

Here are six urgent shifts banks must make now to avoid catastrophic breaches in 2025 and beyond.

1.    Quantum-driven cyber-attacks are a ‘now,’ not a ‘later’ thing

The quantum cybersecurity risk has shifted from the theoretical to the inevitable as cybercriminals get closer to breaking the traditional encryption methods that banks have relied on for decades.

For instance, late last year, a group of Chinese researchers reportedly cracked RSA encryption using a D-Wave quantum computer—a major wake-up call for financial institutions relying on traditional cryptographic defenses.

Once quantum computers become accessible to cybercriminals, traditional encryption methods will be obsolete—and the risk of exposing financial information and sensitive customer data will be at banks’ doorsteps.

  • Implement post-quantum cryptography (PQC): Financial institutions should begin transitioning to quantum-resistant cryptographic methods, leveraging standards such as those published by the National Institute of Standards and Technology (NIST). Quantum-resistant algorithms will act like a futuristic lock, securing data from cyberattacks.

  • Conduct cryptographic inventory & risk assessments: Banks must audit their current encryption framework to identify high-risk financial data. They should prioritize long-term transaction records, customer authentication systems and secure communications to protect against “harvest now, decrypt later” exploitation, where attackers steal encrypted data with the intention of decrypting it once commercial quantum computing becomes viable.

  • Adopt a hybrid encryption strategy: Transitioning to a full-scale PQC presents a significant financial burden. A hybrid approach, combining classical and quantum-resistant algorithms, enables a more gradual and manageable transition.

2.    AI-powered cyberattacks need AI-powered cybersecurity

Cybercriminals can use AI or machine learning algorithms and techniques to launch a range of cyberattacks, such as adaptive phishing scams, deepfake-driven fraud and malware that learns banking security patterns and adapts to bypass detection.

In a recent example, cybercriminals tricked an employee into authorizing $25.6 million in fraudulent transactions by creating AI-generated synthetic voices and video replicas of a senior executive.

This AI-driven social engineering attack bypassed traditional fraud detection methods because the deepfake perfectly mimicked the executive, making it virtually indistinguishable from a real video call. This proves that manual verification processes are no longer enough—organizations must implement AI-powered defense mechanisms to safeguard transactions.

  • Move from basic AI fraud detection to predictive threat intelligence: Banks must be prepared to forecast an attack vector before it escalates. Predictive threat intelligence, powered by AI, can scan global cybercrime trends and monitor hacker forums, underground marketplaces and evolving malware strategies to anticipate fraud techniques before they are deployed. This allows the financial institution to preemptively strengthen defenses before an attack takes place.

  • Implement AI-powered risk assessment platforms: Machine learning algorithms can help banks analyze data, identify patterns and predict potential security threats. For example, Cognizant collaborated with a global bank to implement a machine learning platform that can forecast vulnerabilities and identify suspicious activities by analyzing large datasets. This predictive tool significantly reduced false alarms and enhanced accuracy in threat identification.

  • Move to an AI-powered fraud detection system. Banks should use predictive analytics to identify suspicious transactions and activities in real time. These systems monitor transactions and identify suspicious activities such as a sudden large withdrawal from a rarely used account. The system learns from new data to improve accuracy and reduce false alarms, ensuring genuine transactions aren't mistakenly flagged.

  • Use AI to analyze network traffic patterns. AI can help detect unusual network activities that may indicate a cyber-attack. Picture a traffic cop monitoring the data flow in and out of bank's network. If a sudden spike in data is being sent to an unknown server, the system can alert the bank to investigate further.

3.    Fragmented security systems leave vulnerability gaps that need to be closed

When security tools like firewalls, endpoint security and cloud protection aren’t integrated, it’s all too possible for attackers to exploit these gaps, infiltrating financial networks where banks lack full visibility.

An example is the Bangladesh Bank Heist, where cybercriminals triggered 35 fraudulent transactions, resulting in $101 million in unauthorized transfers. Without a unified security ecosystem, fraudulent transactions could not be immediately detected, allowing large-scale fund transfers before intervention.

  • Implement extended detection and response (XDR). Banks typically use multiple levels of digital security, but these are often not integrated. With XDR, they can integrate their security tools into a cohesive system, like a control room, providing a unified approach to cyberattacks.

  • Use cybersecurity mesh architecture (CSMA). CSMA integrates various security tools and technologies into one cohesive framework, which allows banks to deploy security controls closer to their assets, instead of depending on a centralized security system to respond. This decentralized approach improves response times to threats and reduces widespread breaches.

4.    Automation is the only way to keep up with ransomware

From 2021 to 2024, the share of financial institutions experiencing ransomware attacks increased significantly. In 2024, roughly 65% of financial organizations reported experiencing a ransomware attack, up from 34% in 2021. 

  • Deploy AI-automated incident response and security orchestration, automation and response (SOAR). A SOAR platform reduces the time it takes to address threats and minimizes human error. Picture a fire alarm system that detects fire, activates sprinklers and calls the fire department simultaneously.

    SOAR rapidly detects and contains ransomware threats, preventing encryption and spread by automating security responses. It integrates with security information and event management (SIEM) systems, firewalls and AI analytics to ensure swift, precise threat mitigation. Unlike manual systems, SOAR eliminates delays, reduces alert fatigue and enhances security efficiency by prioritizing critical threats in real time.

5.    Insider threats and credential theft require zero-trust security

Financial systems are increasingly vulnerable to credential theft, where cybercriminals exploit stolen employee credentials to gain unauthorized access. Legacy banking security relies on implicit trust, assuming that users with access are legitimate. However, this approach fails to detect compromised credentials, allowing attackers to operate undetected within sensitive banking systems.

  • Deploy zero-trust security models. The zero-trust model enforces strict authentication policies, verifying identity at every access point rather than assuming internal users are safe. The model employs advanced identification techniques and continuous monitoring to verify authenticity and detect potential threats.

  • Use AI-driven behavioral analytics: AI continuously monitors access patterns, detecting anomalous behavior that signals compromised credentials—even when attackers use valid login details. Imagine banks as a busy marketplace where AI acts like a vigilant security guard who knows the regular customers.

    For instance, we partnered with a leading bank to deploy an AI-driven system combining optical character recognition (OCR) and neural networks to analyze handwritten checks. This solution enabled real-time identification of fraudulent checks, drastically reducing manual efforts and improving operational efficiency, resulting in projected annual savings of $20 million.

6.    Cloud banking introduces new banking cybersecurity challenges that require scalable security

As banks transition to cloud-native infrastructure, traditional cybersecurity approaches struggle to scale against increasingly sophisticated cyber threats. Legacy perimeter-based defenses are vulnerable to misconfigurations, large-scale attacks and rapid threat expansion, making scalability a critical security priority.

  • Implement advanced cloud security infrastructure: Banks need advanced cloud infrastructure and security tools that protect data and applications in a scalable, cost-effective way. These systems need to incorporate cutting-edge security measures, AI-driven automation and real-time analytics to enhance protection against cyber threats. Micro-segmentation ensures that if one security system is compromised, the attack doesn’t spread across the entire banking network.

    For instance, when Indian banks were targeted by hacktivist groups in 2024, the criminals launched distributed denial of service (DDoS) attacks to disrupt online banking services and expose customer data. The banks used their cloud security solutions to absorb and distribute attack traffic, preventing service disruptions. The AI-driven cloud monitoring systems detected unusual traffic spikes, allowing the banks to block malicious requests before they overwhelmed systems.


Banking cybersecurity can’t be a patch—it requires a paradigm shift

The financial institutions that will thrive in 2025 and beyond are not those that react to breaches—they are the ones that anticipate them, adapt in real time and architect security into every layer of their digital infrastructure. This is not just about protecting data—it’s about preserving trust, ensuring operational continuity and safeguarding the very foundation of modern finance. The time for transformation is now.
 

Nageswar Cherukupalli

SVP & BU Head, BCM and Strategic Initiatives

Nageswar Cherukupalli

Nageswar is a Senior Vice President and Head of Banking and Capital Markets. He is a 25-year industry veteran with expertise spanning sales, strategy, consulting, marketing and general management. ​Nagesh is an alumnus of Harvard Business School and has a keen interest in content, culture, and collaboration.

Latest posts

Digital transformation for a next-gen banking revolution

Visit the Banking section of our website.

Know more
Girl watching laptop

Related posts

  • /content/cognizant-dot-com/us/en/insights/insights-blog
  • cognizant:industries/banking
  • cognizant:services/security
  • cognizant:topics/deep-dive
Subscribe for more and stay relevant

The Modern Business newsletter delivers monthly insights to help your business adapt, evolve, and respond—as if on intuition