AI technology is becoming essential for anti-money laundering [AML] processes in financial institutions, helping them comply with regulations and combat financial crime. However, a study by SAS, in collaboration with KPMG, shows that while interest in AI is high, its full implementation remains limited.
AI Adoption in AML
The study, based on a survey of 850 AML specialists, found that AI and machine learning [ML] adoption is still modest:
- 18% have AI/ML solutions in production.
- 18% are piloting AI/ML solutions.
- 25% plan to adopt AI/ML in the next 12-18 months.
- 40% have no plans to use AI/ML.
Generative AI [GenAI] is also gaining attention, though cautiously. About 10% of respondents are piloting GenAI, and 35% are exploring it, but 55% have no plans for adoption.
Regulatory Concerns
The study highlights changing regulatory attitudes. In 2021, 66% of respondents believed regulators encouraged AI use, but that number has dropped to 51%. Meanwhile, those who view regulators as “resistant to change” more than doubled from 6% to 13%.
Benefits of AI in AML
AI and ML have proven effective in specific AML functions, especially where large data volumes need to be processed. Key benefits include:
- Automating alerts from transaction monitoring.
- Improving risk assessments and suspicious activity reporting.
- Reducing false positives in AML checks.
Challenges to AI Adoption
Despite its benefits, AI adoption in AML faces obstacles. In 2021, budget constraints [39%] were the biggest hurdle, but in this survey, the lack of regulatory imperative [37%] has overtaken budget concerns [34%]. Additionally, the lack of available AI skills, once a major barrier, has dropped to 11%.
Priorities for AI Implementation
Reducing false positives in AML surveillance is a growing priority, with 38% of experts citing it as their main focus [an 8% increase from 2021]. Other priorities include:
- Automating data enrichment for investigations [25%].
- Detecting new risks using advanced modeling [23%].
- Enhancing customer segmentation for behavioral analysis [13%].
AI Technologies Making an Impact
Among AI technologies, machine learning [ML] dominates, with 58% ranking it as the most impactful tool. Robotic process automation [RPA] follows at 28%, while natural language processing [NLP] remains underutilized [14%], despite its potential to detect early warning signs in AML cases.
Moving Forward with AI in AML
To maximize AI’s potential, financial institutions must integrate their data sources, teams, and technology. The survey found that 86% of organizations have started integrating AML, fraud, and information security functions. Those that fully integrate AI-driven AML processes will gain a competitive advantage over firms that hesitate.
Ultimately, firms that proactively implement AI while aligning with governance and compliance frameworks will be better positioned to fight financial crime, navigate evolving regulatory landscapes and champion anti-money laundering.
