Report: Fraud Teams Are Growing Despite Widespread AI Use
A new report from SEON based on a survey of over 1,000 fraud and compliance leaders finds that while AI use is nearly universal, fraud prevention teams are still increasing in headcount. The data indicates that companies are expanding their budgets and staffing for fraud and AML (Anti-Money Laundering) roles. This growth occurs alongside the challenges of managing fragmented systems.
- According to the SEON report, which surveyed 1,010 fraud, risk, and compliance leaders, 94% plan to add at least one full-time hire in 2026, an increase from 88% in 2025. Additionally, 83% of these leaders expect their fraud and AML budgets to increase. - A significant challenge limiting AI's effectiveness is system fragmentation; while 95% of organizations claim some integration between fraud and AML systems, only 47% operate with fully integrated workflows. This fragmentation is a key reason why 80% of leaders find it difficult to get a unified view of their data. - The use of AI by criminals is a primary driver for increased fraud team staffing, with AI-driven scams seeing a 1,210% increase in 2025. These sophisticated attacks include deepfake video impersonation, AI voice cloning, and advanced business email compromise (BEC) schemes that traditional defenses struggle to detect. - While nearly all organizations (98%) have integrated AI into their fraud and AML workflows, the technology has exposed the sheer volume of work rather than reducing it. AI excels at tasks like transaction monitoring, which is its top use case (30%), but human oversight remains critical to interpret nuances and establish intent, which AI currently cannot do. - The global market for fraud detection and prevention is projected to reach $67.12 billion in 2026, reflecting a compound annual growth rate of 17.5%. This market growth is fueled by the widespread adoption of digital payments and the necessity for real-time authentication. - Despite the high adoption of AI for defense, many institutions feel exposed, with 99% acknowledging flaws in their detection capabilities. Key issues include failures in sanctions screening (23%), siloed datasets (22%), and a lack of real-time visibility (21%). - Implementation of new fraud prevention vendor solutions is often a lengthy process, which prolongs fraud exposure. Only 10% of companies go live in under two weeks, while 38% take 1-3 months, and 24% require more than four months. - Criminals are increasingly using generative AI to create "DIY" fraud tools, including deepfake content, malicious code for scam websites, and highly convincing phishing emails that now deceive 60% of recipients. This has contributed to a 1,265% surge in phishing attacks linked to generative AI trends in the past year.
