Wednesday, September 10, 2025

Unlocking Model Potential:
Advanced Tuning, AI, and Auditing for Compliance
Presented by AIBACP and Cygnus Compliance

Half-day, In-Person Symposium
12:30 p.m. EDT – Guest Arrival / Sign-in
1:00 p.m. – 4:00 p.m. – Speaker Presentations

Refreshments will be served

New York City Bar Association
42 West 44th Street, New York, NY

REGISTER TODAY!

Curious how advanced model optimization, artificial intelligence, and governance intersect to transform AML, sanctions,
and regulatory compliance? Join us as Cygnus Compliance and leading voices from the international banking
community convene for a dynamic, three-panel program tackling the most pressing opportunities and risks facing
financial institutions today.

Panel I: Tuning with Precision: Advanced AML & Sanctions Model Optimization (Speakers TBA)
Key Themes:

• Unlocking model potential through back-testing, threshold calibration, and intelligent tuning
• Continuous model monitoring: KPIs, false positive reduction, and regulatory defensibility
• Fulsome documentation, audit trails, and examiner-ready validation packages
• Integration of enriched data sources (i.e. ISO 20022 fields) into tuning frameworks

Takeaway: Actionable techniques to maintain AML effectiveness while enhancing efficiency and governance

Panel Il: Smarter Compliance: The Role of AI in AML, Sanctions, and Investigations (Speakers TBA)
Key Themes:

• Leveraging GenAI and LLMs for intelligent alert triage, SAR drafting, and investigator workflows
• Risks and controls: mitigating hallucinations, bias, and explainability issues / Black box problems
• AI model governance in regulated environments

Takeaway: Practical AI adoption strategies that reduce friction and boost accuracy across the compliance lifecycle

Panel lll: Auditing AI at Financial Institutions – Risk, Assurance & Governance (Speakers TBA)
Key Themes:

• Understanding hallucination and drift: how LLMs can produce misleading outputs and how auditors can assess robustness and reliability
• Evaluating governance and control frameworks: ensuring AI systems are designed and operated with accountability, transparency, and fairness
• Auditing data sourcing and model integrity: identifying risks related to bias, data lineage, and third-party AI tools
• Testing explainability and monitoring procedures: assessing whether internal AI systems support decision-useful, interpretable outcomes
• Benchmarking against emerging standards: comparing internal practices to leading frameworks and evolving regulatory expectations

Takeaway: Why internal audit functions must evolve to provide independent, evidence-based assurance over AI systems – and how doing so can build trust, mitigate risk, and support responsible innovation.

Members may register as follows: sign-in on the website https://www.aibacp.org/ , select “Upcoming Events”
Non-members who wish to attend may apply for membership on the website
CPE Credit: 3 CPE credits will be awarded to members attending the in-person seminar

Agenda to follow