Ayasdi had the honor of sponsoring the Association of Certified Anti-Money Laundering Specialists (ACAMS) conference in Las Vegas in September. This is the flagship event for ACAMS and it showed in the audience. Present were the heads of Anti-Money Laundering (AML) programs at every major financial institution, consultants, and even regulators.
We thought it would be beneficial to share the top five key takeaways that we uncovered at the conference.
There is a clear need for Artificial Intelligence (AI) in AML
In talking to customers, a common theme emerged – that AI was the future of AML. People openly spoke about the limitations of archaic, rule-based processes that were ineffective in catching money launders as well as inefficient operationally. They also discussed the need for more advanced technologies that could learn – constantly adapting to changing money laundering patterns.
Artificial Intelligence in AML is still new and buyer beware
While there is broad recognition of the future impact of AI on AML, there is some concern about the here and now. As with any technology, there were a lot of vendors touting the title of being an “Artificial Intelligence” company for AML. We think buyers need to beware and do their diligence. It doesn’t take too long to find that these “purported” AI solutions are not production ready and no customer has moved past the pilot phase. Given the amount of claims on the show floor, we were surprised to find that Ayasdi was the only company discussing production examples of their AI for AML solution.
There is limited differentiation in AI vendors for AML
In our conversations with other vendors, we found that the majority of them were focusing primarily on the Alert Investigation part of the process. Though this is a core element of any AML process, these vendors are providing more of a data aggregation solution than an AI solution. This was somewhat common.
Figure out how to be seamless
AML systems and processes are highly regulated. While AI offers tremendous benefits, it cannot be disruptive. AI’s ability to provide order of magnitude improvement in performance provides a range of attack points. For instance, at Ayasdi, our Intelligent Segmentation capability can fit into existing AML processes in a frictionless way to provide a more granular and justifiable customer segmentation. This Intelligent Segmentation, which can be deployed in a matter of weeks, can help tune scenario thresholds in a way to reduce the number of false positives, while maintaining the same level of risk mitigation.
Find the lever point
We found that many customers were discussing 4 primary ways that AI would augment their existing AML processes and systems. These represent lever points where the ROI can be massive:
- Customer Segmentation: the ability to use AI to more granularly segment a customer population in an unsupervised way to feed into the existing threshold tuning process
- Alert-Auto-Dispositioning: automatically categorize and route alerts to the right alert investigator level, ensuring a reduction of alerts for each investigator level and a streamlined alert investigation process
- Dynamic Typologies: identify anomalies in customer behaviors in relation to other similar customers, rather than create rule-based typologies
- Customer Risk Ratings: leverage AI and publically available data to create more comprehensive and accurate customer risk ratings
Again, this an extraordinary time for the AML space. The traditional methods have proven ineffective in the world of heightened regulatory scrutiny and massive fines. At Ayasdi, we believe that the financial services industry has an incredible opportunity to incorporate AI into their existing AML processes and systems. Our Intelligent Segmentation has transformed the false positive challenge for banks and we are prepping the launch of two additional element Intelligent Alerts and Intelligent Typologies. Stay tuned….
To dig a little deeper, click here or reach out to us for a demo.
Ayasdi had the honor of sponsoring the Association of Certified Anti-Money Laundering Specialists (ACAMS) conference in Las Vegas in September. This is the flagship event for ACAMS and it showed in the audience. Present were the heads of Anti-Money Laundering (AML) programs at every major financial institution, consultants, and even regulators.
We thought it would be beneficial to share the top five key takeaways that we uncovered at the conference.
There is a clear need for Artificial Intelligence (AI) in AML
In talking to customers, a common theme emerged – that AI was the future of AML. People openly spoke about the limitations of archaic, rule-based processes that were ineffective in catching money launders as well as inefficient operationally. They also discussed the need for more advanced technologies that could learn – constantly adapting to changing money laundering patterns.
Artificial Intelligence in AML is still new and buyer beware
While there is broad recognition of the future impact of AI on AML, there is some concern about the here and now. As with any technology, there were a lot of vendors touting the title of being an “Artificial Intelligence” company for AML. We think buyers need to beware and do their diligence. It doesn’t take too long to find that these “purported” AI solutions are not production ready and no customer has moved past the pilot phase. Given the amount of claims on the show floor, we were surprised to find that Ayasdi was the only company discussing production examples of their AI for AML solution.
There is limited differentiation in AI vendors for AML
In our conversations with other vendors, we found that the majority of them were focusing primarily on the Alert Investigation part of the process. Though this is a core element of any AML process, these vendors are providing more of a data aggregation solution than an AI solution. This was somewhat common.
Figure out how to be seamless
AML systems and processes are highly regulated. While AI offers tremendous benefits, it cannot be disruptive. AI’s ability to provide order of magnitude improvement in performance provides a range of attack points. For instance, at Ayasdi, our Intelligent Segmentation capability can fit into existing AML processes in a frictionless way to provide a more granular and justifiable customer segmentation. This Intelligent Segmentation, which can be deployed in a matter of weeks, can help tune scenario thresholds in a way to reduce the number of false positives, while maintaining the same level of risk mitigation.
Find the lever point
We found that many customers were discussing 4 primary ways that AI would augment their existing AML processes and systems. These represent lever points where the ROI can be massive:
- Customer Segmentation: the ability to use AI to more granularly segment a customer population in an unsupervised way to feed into the existing threshold tuning process
- Alert-Auto-Dispositioning: automatically categorize and route alerts to the right alert investigator level, ensuring a reduction of alerts for each investigator level and a streamlined alert investigation process
- Dynamic Typologies: identify anomalies in customer behaviors in relation to other similar customers, rather than create rule-based typologies
- Customer Risk Ratings: leverage AI and publically available data to create more comprehensive and accurate customer risk ratings
Again, this an extraordinary time for the AML space. The traditional methods have proven ineffective in the world of heightened regulatory scrutiny and massive fines. At Ayasdi, we believe that the financial services industry has an incredible opportunity to incorporate AI into their existing AML processes and systems. Our Intelligent Segmentation has transformed the false positive challenge for banks and we are prepping the launch of two additional element Intelligent Alerts and Intelligent Typologies. Stay tuned….
To dig a little deeper, click here or reach out to us for a demo.
