The PwC 2025 Global Digital Trust Insights Report found that nearly seven in 10 (67%) security specialists say Gen AI has expanded the payments risk surface. To combat fraud attacks using new scale and quality made possible by Gen AI, the industry is uniting.

Enter the concept of "fraud fusion" – an innovative approach to cybersecurity that breaks down traditional silos between cyber and fraud teams. As digital threats get faster and more convincing using AI and automation, organizations are finding that conventional security measures are no longer adequate. It’s triggering a fascinating development.

As Bottomline Chief Information Security Officer (CISO) Chirag Patel explains in this Q&A, fraud fusion is bringing together specialists from different disciplines, and even different companies, to join forces, sharing intelligence and coordinating actions to harden defenses against modern AI threats. Patel notes how this strategy is supercharging cybersecurity, and why it's indispensable in a payments landscape now dominated by AI.

 

Q: Chirag, what is “fraud fusion”, and how is it taking shape?

Chirag Patel: Fraud fusion is about [in-house departments or different companies] bringing different elements together to create something stronger and more effective than they could achieve on their own. This means cyber and fraud teams collaborating in a shared environment, using sharing intelligence, tools, and strategies rather than operating in silos. In the case of actual fraud fusion centers, think of these as Mission Control Centers for digital threats. Just as NASA brings together specialists from different disciplines to solve complex problems in space missions, fraud fusion centers unite experts from cyber, fraud, and risk domains to tackle sophisticated digital threats. This approach is gaining traction because attackers don't think in terms of organizational silos. Fraudsters look for any vulnerability to exploit. Our defense mechanisms need to match this integrated approach.

 

Q: How does a fraud fusion ‘center’ operate, in practice?

Chirag Patel: In the case of physical facilities, picture a room where experts with different specialities work together, with large screens displaying both cyber and fraud monitoring data. This physical proximity facilitates immediate collaboration when an incident occurs. What makes these centers particularly effective is the breakdown of traditional barriers between departments. In conventional setups, if the fraud team detects suspicious activity, they might need to file a ticket with the cyber team and wait for a response. In a fusion center, they simply turn to their colleague at the next desk and address the issue immediately. This "War Room" mentality is crucial because speed is essential when responding to attacks. Fusion greatly reduces response time and improves coordination, which can make the difference between a minor incident and a major breach.

However, in today’s virtual work environment, this same principle can still be achieved remotely. What’s most important is to build trust and collaboration between IT, security, and fraud departments.

 

Q: Which industries are leading the adoption of fraud fusion models?

Chirag Patel: Industries facing the highest risk of financial and reputational damage are at the forefront, as we would expect. Banking, FinTech, and payments sectors are leading the charge because they're prime fraud targets. We're also seeing significant adoption in healthcare, telecommunications, and e-commerce. It’s catching on in any industry where valuable data or financial transactions are at stake. According to Gartner, the number of organizations adopting a full fraud fusion model is expected to quadruple by 2028. In the early stages, we typically see cyber operation centers and fraud response teams working together as the first line of defense. As organizations mature, collaboration expands beyond company boundaries. More advanced organizations continually interact with federal agencies, law enforcement, and even other companies in the same industry to share threat intelligence, recognizing the value in collective defenses.

 

Q: What makes the fusion approach innovative compared to conventional cybersecurity?

Chirag Patel: The innovation lies in several key areas. First, there's the tech crossover. Tools originally built for cybersecurity can now fight fraud, and vice versa. Second, the fusion model enables a "delayed defense" approach with multiple layers of protection. If a threat actor bypasses one security control, they'll encounter another. Fusion centers help identify gaps between cyber and fraud controls that might otherwise go unnoticed. Third, the human element is transformed. By bringing together specialists with different expertise, fusion centers create an environment where creative problem-solving thrives. A fraud analyst might notice patterns that a cyber specialist wouldn't, and vice versa. Fourth, the fusion approach enables a holistic view of the risk landscape, connecting dots between what might appear to be separate incidents but are, in fact, part of a coordinated attack. Finally, response time is dramatically improved. [Some large organizations] have reported significant reductions in fraud losses after implementing fusion centers, primarily due to faster detection and response times.

 

Q: How are AI and automation making cybercrime worse?

Chirag Patel:  The cybercrime and fraud landscape continues to evolve at an alarming rate, fueled by the rapid evolution of the digital landscape and advancements in technology. Fraud patterns such as Phishing, Business Email Compromise, identity theft, and Bot attacks are still common. What has changed is how these attacks are performed. We're seeing a significant increase in both the scale and sophistication of attacks, with about 62% of businesses experiencing cybercrimes powered by AI-driven techniques. These technologies enable attackers to operate on an unprecedented scale. Phishing campaigns that once required substantial manual effort can now be automated and personalized using AI. Identity theft has become so prevalent that the FTC receives reports every 22 seconds. Business email compromise [BEC] has resulted in staggering losses of over $55 billion in the past decade. What makes AI-powered attacks particularly dangerous is their ability to learn and adapt. Traditional security measures that rely on known signatures or patterns are growing ineffective against these dynamic threats. Attackers are using AI to identify vulnerabilities, craft convincing social engineering campaigns, and even develop malware that can evade detection.

 

Q: How does fraud fusion specifically confront these new AI-powered attacks?

Chirag Patel: Just as attackers use AI to scale their operations, we need to deploy AI-powered defenses that can analyze vast amounts of data, identify subtle patterns, and respond at machine speed. The combination of cyber and fraud data provides a richer dataset for these AI systems to learn from, making them more effective at detecting anomalies. The human expertise in fusion centers is equally important. AI systems excel at processing data and identifying patterns, but human analysts bring contextual understanding and intuition that machines lack. In a fusion environment, these human experts can collaborate to interpret AI findings and make nuanced decisions about potential threats. Perhaps most importantly, the fusion approach enables organizations to be proactive rather than reactive. By combining threat intelligence from multiple sources and domains, security teams can anticipate emerging threats and strengthen defenses before attacks occur.

 

Q: What metrics suggest that the fraud fusion approach is improving outcomes?

Chirag Patel: The effectiveness of fraud fusion is evident in several key performance indicators. First is the reduction in financial losses. Companies implementing fraud fusion have reported decreases in fraud losses by as much as 60-70%. Another important metric is mean time to detect (MTTD) and mean time to respond (MTTR), which typically decrease by 30-50% compared to traditional approaches. False positive rates also tend to improve with fusion. By combining data and insights from multiple domains, organizations can better distinguish between genuine threats and benign anomalies. We're also seeing improvements in regulatory compliance metrics, as the comprehensive visibility provided by fraud fusion helps ensure nothing falls through the cracks. Perhaps the most telling metric is the evolution of attack patterns. When organizations implement fraud fusion, they often observe that attackers shift to softer targets. This suggests that the fusion approach is creating a more formidable defense that adversaries prefer to avoid.

 

Q: How would one go about starting a cyber-fraud fusion?

It doesn’t have to be a complicated or massive undertaking. Here are some simple ways to get started with fraud fusion:

  • Get leadership on board. You’ll need their support to make this work.
  • Start small. Encourage collaboration between your cyber and fraud teams; share insights, compare notes, and build trust.
  • Let it grow organically. Once people see the value, the collaboration will naturally deepen over time.