Jack Henry & Associates, Inc. · NASDAQ:JKHY
Jack Henry & Associates — Credit Union Strategy Summit 2024: Session 1
38 slides from Jack Henry & Associates's conference presentation, Jul 23, 2024
Slide 1: AI at the Forefront: Revolutionizing Fraud Detection and Mitigation
Credit Union Strategy Summit: Session 1 AI at the Forefront: Revolutionizing Fraud Detection and Mitigation Navigate the future, Empower your Organization 07/23/2024
Slide 2: Contents
Current State of AI 04 AI Strategy & Principles 07 10 contents Understanding the Risks Shaping your Strategy 12 Financial Crimes 13 Defender & AlertCenter Poll 36 Q&A 37 2
Slide 3: About the presenter: Charlie Wright
about the presenter Charlie Wright Chief Risk Officer 405.834.0464 ctwright@jackhenry.com 3
Slide 4: The Current State of AI
The Current State of AI
Slide 5: The Current State of AI
The Current State of Al jack henry
Slide 6: Financial Services is at the Forefront of the AI Transformation
Financial Services is at the Forefront of the AI Transformation Technology Convergence Create Your Competitive Advantage Establish Your Strategy 6
Slide 7: AI Strategy
AI Strategy 1. Governance and risk management 2. Engineering architecture 3. Tool and model management 4. Data Management 5. Use Case Inventory and development 6. Third Party Partners 7. People, culture and change management 8. Communication 7
Slide 8: Establish AI Principles
Establish AI Principles Fairness and Equality Compliance Promote fair, just, and equitable Compliant with all applicable laws, treatment and avoid discrimination regulations, and standards. and bias. Transparency Privacy and Data Protection AI use cases should be clearly Respect privacy, protect dat
Slide 9: Have You Defined AI?
Have You Defined AI? What it is What it is not • Machine Learning • Robotic Process Automation • Computers analyzes existing data • Automates manual, repetitive, and rule- and learns to identify patterns based tasks • Generative AI • Learns statistical patterns and then generates new data based on t
Slide 10: Understand the Risks Associated with Specific Models
Understand the Risks Associated with Specific Models Market Fragmented by Functionality Google OpenAI Microsoft Free Gemini ChatGPT 3.5 MS Copilot ChatGPT - 4o Subscription Gemini 1.0 Pro ChatGPT 4.0 MS Copilot Pro Gemini 1.5 Pro ChatGPT Plus MS for Office 365 Gemini 1.0 Ultra 10
Slide 11: Use Case Approval Processes
Use Case Approval Processes Organize into four categories of use cases General Business Software Third Party Productivity Use Development Integration 11
Slide 12: What is Your Institution's AI Strategy Framework?
What is Your Institution’s AI Strategy Framework? Institution’s AI Principles Strategic Plan Risk Management Compliance Data Privacy Security Use Case Approval Process Communication Training and Professional Development Tools/Technology Model Management Data Management 12
Slide 13: Financial Crimes Defender
Financial Crimes Defender
Slide 14: About the presenter: Rene Perez
about the presenter Rene Perez National Director of Sales - Financial Crimes Solutions 205.981.1980 rperez@jackhenry.com 14
Slide 15: Answer the Poll on your screen
Answer the Poll on your screen
Slide 16: Fraudemic
Fraudemic $151B $4.23 Estimated total amount of fraud in the US The amount it costs a Financial Institution Payments $102.6B for every fraud dollar Check $21B Scams $13.9B Credit card $13.6B Source: American Banker Source: LexisNexis 16
Slide 17: Social Media
Social Media 17
Slide 18: Phone Tree
Phone Tree 18
Slide 19: @ wee.
@ wee.
Slide 20: Sample driver's license
01/06/1985 ‘ 01/06/2028 01/06/1985 BENDER WELLS HIRAM 1050 MAIN WEST ENFIELD, ME 44934-426 12/05/2022 M 130 Ib GRY BRO c NONE NONE oN DRIVER'S LICENSE N M Si paras 9702079 Rad aia : Ad a
Slide 21: 21
21
Slide 22: Synthetic ID Account Drop
Synthetic ID Account Drop • This was a sample of a drop 22
Slide 23: Legacy Systems Issue
Legacy Systems Issue • Yellow Hammer, Verafin, Abrigo, Nice-Actimize • All started as BSA systems • Not designed for real-time • They are piecemeal fraud add on’s that are not highly effective 23
Slide 24: Fraud Industry Problem
Fraud Industry Problem 24
Slide 25: How to Disrupt Fraudsters
e How to Disrupt Fraudsters 25
Slide 26: The Defender Platform
The Defender Platform Mobile Online Branch Call Center Card/ATM eCommerce Digital Channels Core Payment Channels Financial Crimes Defender Cloud Based True Real-time Digital First AI/ML Open Architecture Investigations Mitigating Loss Reporting 26
Slide 27: Decisioning Engine
Decisioning Engine • Account Profiles • Receiving Profiles 27
Slide 28: Fraudster BoA Acc 1234
Fraudster BoA Acc 1234 Acc 8989 - Acc 2345 - Acc 5678 - Acc 9876 - Acc 1212 - $5,000 $1,000 ACH $1,000 Zelle $5,000 Wire $1,500 RTP Checks
Slide 29: Decisioning Engine
Decisioning Engine • Account Profiles • Receiving Profiles • Digital Profiles • Velocities • Behavior • Machine Learning • Consortium • National • FI Specific 29
Slide 30: Consortium Modeling
Consortium Modeling • Centralized Non-PII Data • Shared Detection • Data Science • Machine Learning Models • Champion • Challenger • Constant Updates • Ability to Ingest Additional Data 30
Slide 31: What is Jack Henry Doing?
What is Jack Henry Doing? • Continue to build out Financial Crimes Defender • Consortium Data is key • Partnerships are Key for Additional Data • Feedzai • American Banker’s Association (ABA) • Sardine • Mastercard • Google • Dow Jones • U.S. Federal Reserve 31
Slide 32: Top Banks – Zelle Fraud Detection Rates
Top 20 Banks – Zelle Fraud Detection Rates 53% to 83%* Rate has direct relationship to FI’s risk tolerance vs Friction Models have been fine tuned for 5+ years *These are estimates based on multiple different data sources as information is not public 32
Slide 33: Jack Henry's Zelle Model Performance Overview
Jack Henry’s Zelle Model Performance Overview • Model based on limited historical data set containing 1272 FI’s • Model Training • 1,923,017 transactions • This data does not include two key • Model Validation data points: • 1,054,966 transactions • IP Address • Machine ID Transaction Money • Estima
Slide 34: So what?
So what? 34
Slide 35: So What?
So What? • Fraudsters are organized • Fraudsters are more sophisticated than people realize • Single point fraud solutions are a false sense of security • Consortium data across all channels are key for the future 35
Slide 36: Answer the Poll on your screen
Answer the Poll on your screen
Slide 37: Q&A
Q&A
Slide 38: © 2023 Jack Henry & Associates, Inc. ®
© 2023 Jack Henry & Associates, Inc.®
Jack Henry & Associates — Credit Union Strategy Summit 2024: Session 1 on 6ix