Artificial intelligence is not simply giving fraudsters new tools. It is changing the scale at which fraud can operate.
That’s one of the central themes of a new ID Talk conversation with Maxine Most, founder and CEO of Acuity Market Intelligence, creator of the PRISM Project, and organizer of the Deepfake Summit. Speaking with ID Talk hosts Alex Perala and Tony Bitzionis ahead of the next Deepfake Summit in Arlington, Virginia, Most argued that AI is making sophisticated fraud techniques dramatically easier to automate and deploy.
“Basically what AI has done is industrialized fraud threats,” Most said. “It means you don’t need sophisticated technology to go apply for 10,000 credit cards in an hour.”
For Most, that ability to operate at scale is one of the most important changes taking place across the fraud landscape. Deepfake video and audio may attract much of the attention around generative AI, but the broader problem extends to synthetic identities, injection attacks, employment fraud, phishing, forged documents, and other forms of impersonation.
The result is a threat environment that increasingly cuts across departments and industries.
Most pointed to financial services as one area where organizations have already had to confront substantial fraud losses, but she argued that the issue is expanding well beyond traditional fraud teams. AI-enabled attacks can affect intellectual property, hiring, operations, cybersecurity, identity verification, and other parts of an organization.
“We’re getting to this tipping point where organizations are saying, ‘Oh, we really need to do something about this,’” she said.
That raises a harder question: what exactly should organizations do?
Most argued that the answer is unlikely to come from any single product, department, or policy intervention. Instead, organizations may need to rethink how they establish trust across the entire identity lifecycle.
“Fundamentally we’re shifting the security model from trying to put boxes around things with locks on them to who are you? How do you prove who you are? What do you have access to?” she said.
That shift is also helping to shape the Deepfake Summit, which Most launched earlier this year with an initial event in Houston. The next Summit takes place September 1 in Arlington, Virginia, and expands the format from a half-day pilot into a full-day program.
Most described the Summit as an attempt to move away from the conventional conference model of PowerPoint presentations and one-way panel discussions. Instead, the event is designed around conversation between participants from different sectors, including government, financial services, technology, privacy, fraud, policy, healthcare, and emergency management.
The event operates under the Chatham House Rule, allowing participants to speak freely without having individual remarks attributed to them. Most also records the broader discussion and uses it to produce a post-event intelligence report, with the findings from one Summit feeding into the agenda for the next.
The goal, she said, is to create an ongoing conversation rather than a series of isolated conference sessions.
That approach reflects what Most sees as the fundamentally cross-disciplinary nature of the problem. Fraud detection, identity verification, privacy, governance, cybersecurity, and AI policy can no longer be treated as separate issues when the same technologies increasingly affect all of them.
One concept running through the upcoming Summit is what Most calls “resilient trust.” Rather than treating security, privacy, compliance, and fraud prevention as competing priorities, the idea is to think of them as components of the same problem.
The Arlington agenda will also address agentic AI and the growing need to establish trust not only between people, but between humans, machines, and autonomous software agents. Other discussions will focus on fraud across the identity lifecycle, governance, real-world deployments, privacy and data ownership, and the broader question of what trust infrastructure needs to look like in an AI-driven environment.
Underlying those discussions is an increasingly basic problem: determining whether the things people encounter online are authentic at all.
“We live in a world where we cannot trust what we see with our own eyes,” Most said.
And she stressed that this issue goes far beyond realistic AI-generated video.
“It’s documents, it’s credentials, it’s utility bills, it’s pictures of your grandma, it’s everything.”
Most warned that the consequences could eventually extend beyond fraud losses or cybersecurity incidents. If individuals and institutions lose the ability to establish authenticity, she argued, basic social and commercial interactions become much harder to sustain.
“If we don’t figure out how to solve for this, it leads to societal breakdown,” she said. “How do you trust anybody? How do you do business with anybody?”
The full ID Talk episode also explores the upcoming Deepfake Summit agenda, Most’s latest PRISM Project research into financial services, and why she believes collaboration across traditional organizational and industry boundaries will be essential to keeping pace with AI-enabled threats. Listen wherever you get your podcasts, or via the web player below.
–
By the ID Tech Editorial Team
