Banks bet big on AI to fight financial crime, reaping millions in savings

By Puja Sharma

Banks Accelerate AI Adoption, Realising Millions in Savings as Agentic and Generative AI Shape the Future of Financial Crime Compliance

Banks globally are realising measurable impact from artificial intelligence (AI) across financial crime and compliance (FCC) programs — from immediate cost savings to transformative efficiency gains — according to new research from Hawk, the leading provider of AI-powered fraud prevention, anti-money laundering (AML), and screening solutions, in partnership with Chartis Research.

The new report, AI in Financial Crime and Compliance: Charting the Path from Pilot to Maturity (Banking Edition), highlights how financial institutions are moving beyond experimentation to full-scale execution. As next-generation technologies such as agentic AI and generative AI (Gen AI) advance, banks are already seeing results: 71% report measurable cost savings, 57% report improved detection accuracy, and 53% expect annual savings exceeding $5 million within two years of adoption. These findings underscore how banks are embracing AI not just for innovation’s sake but as a proven driver of efficiency, accuracy, and regulatory readiness.

Tobias Schweiger, CEO of Hawk, said: “This research confirms that the banking sector is at a pivotal moment with AI. The realised value is proving to be significant, and now the challenge is scaling that value with a holistic application of different AI technologies. Balancing the strengths of machine learning, GenAI, and Agentic AI for different uses — from high-volume data processing to complex investigations — is key to sustaining momentum and keeping AI costs under control.” The research shows that realised AI benefits often exceed early expectations. While most banks initially prioritised risk coverage and scalability, institutions now cite improvements in investigative quality, decision support, and operational throughput as top outcomes.

Agentic and generative AI are poised to redefine compliance agility. Gen AI is already delivering value in data and document processing, investigative guidance, and model development, while agentic AI is beginning to transform how teams conduct investigations, draft SAR narratives, and synthesise contextual intelligence. Together, they represent the next major leap in augmenting human decision-making and enhancing the explainability of complex compliance processes.

Sean O’Malley, Research Director at Chartis Research, added: “Banks aren’t asking if AI works anymore. Instead, they’re focusing on how to scale it. With regulators becoming more open and the benefits increasingly tangible, we predict that 2026 will mark a decisive acceleration in AI maturity across financial crime and compliance.” This accelerating momentum signals a maturing approach to AI strategy — one focused on embedding explainable, scalable, and responsible intelligence at the core of compliance operations.

Key findings:

• Around 62% of banks plan further AI adoption in FCC within the next 12 months.
• Over 86% see machine learning as having the greatest impact, followed by Gen AI (83%) and Agentic/NLP applications (76%).
• About 82% expect overall AI investment to rise by more than 25% in the next year.
• Over half (60%) believe regulators will become increasingly supportive of AI adoption by 2028.