AI-Enabled Fraud: Expert Insights from Blank Rome LLP Summit

It was a great day at the Blank Rome LLP and FTI Consulting Financial Crimes and Corporate Investigations Summit discussing hot topics in white collar defense and investigations. The morning started with a fantastic set of experts discussing cutting edge developments on the AI and the New Age of Fraud panel. Thank you to Nicole Costaldo, Philip Yannella, Melanie Harris, and moderator Chris Dysard for the thoughtful discussion and for your valuable insights. As Melanie explained – AI impacts the speed, precision, and scale of fraud. Schemes can be hyper personalized at scale. Voices can be cloned with just seconds of recorded sound. To put it succinctly: “your gut instinct is no longer enough.” Nicole taught us that AI-enabled fraud is projected to cost US companies $40 billion in annual losses by 2027. She recommended simulating real threats that have occurred to train employees on the latest risks. Nicole reminded us to review our control frameworks and to think about: - Biometrics - Multifactor authentication - Domain protocols - Third-Party testing to ensure systems are up to date Phil informed us of the latest from state lawmakers – many states, like PA, are passing new AI laws that make it a crime to use deepfakes to commit fraud. Nicole’s recommendation is to fight technology with technology. Machine learning is extremely powerful for detection and finding false positives. The panelists discussed how Anthropic, the developer of the AI model Claude, has published several reports detailing how malicious actors have misused the AI for various fraud and cybercrime campaigns. AI has been used for reconnaissance, identifying vulnerabilities, and generating exploit code. Melanie closed the panel with this thoughtful insight – AI is not inherently good or bad, it depends on how it is trained. The room was captivated as the panelists shared their boots on the ground experience and taught us how to be thoughtful in our approach to managing and combatting against AI-enabled fraud.