Why Legacy Identity Verification Can’t Stop AI-Enabled Fraud

Verification frameworks were built for a different era – one defined by credential theft, static signals and manual review.

Today’s threat landscape is different. 

AI-enabled fraud including deepfakes, synthetic identities and automated agent attacks now operate at machine speed, exploiting weaknesses in onboarding and compliance systems.

The result is clear: verification systems built for yesterday’s fraud no longer work – and the costs are resounding across the digital economy.

According to a joint Trulioo and PYMNTS report, 96% of firms feel confident in detecting harmful bots, yet 59% of them struggle with synthetic identity fraud and generative AI fraud. These gaps have contributed to an estimated $95 billion in annual losses across global enterprises, especially as capabilities of automated fraud have accelerated over the past 12 – 18 months.

“With generative AI fraud, manual analysts can no longer tell legitimate from fraudulent with the human eye. AI-enabled fraud is just that sophisticated and powerful.” said Trulioo Chief Product Officer Zac Cohen. “You need to be able to have access to the latest, most sophisticated and innovative tools to protect your system now.”

In a recent PYMNTS roundtable, Cohen joined Will Fitzgerald, vice-president of global fraud and financial crimes at WEX, to address a critical question: In a new reality of AI-enabled deepfakes, synthetic identities and automated agent attacks, how can firms rethink legacy identity frameworks – and how can stronger digital identity stop fraud while lifting conversion and revenue? 

How Compliance Turns From Cost Centre to Revenue Unlock

According to the report, 52.3% of companies are seeing increased bot traffic, while 52.9% report onboarding drop-off, contributing to an average 3.1% revenue loss. When legacy systems fail to distinguish between fraudsters and legitimate users, conversion drops, acquisition slows and operational costs rise, undermining fraud operational efficiency.

Yet even as generative AI fraud and outdated onboarding flows accelerate those major losses, compliance and digital identity systems are still often treated as cost centers rather than growth drivers.

By investing in modern, adaptive AI fraud prevention tools, enterprises can reduce fraud losses while improving onboarding for legitimate users. More precise identity verification allows low-risk users to move through faster while applying intelligent friction to higher-risk activity.

It’s here that compliance moves from a cost center to what Cohen calls a “revenue unlock” – a way that global enterprises can directly convert legitimate users by reducing false positives,and curb fraud losses by applying intelligent friction.

“There are a lot of good opportunities here for professionals and for experts to push the boundary to get better compliance, better risk screening, and ultimately better revenue.” Cohen said.

Reputation compounds the impact. Strong compliance and smooth onboarding build trust and support growth. Excessive friction has the opposite effect, driving legitimate users away and eroding brand value.

“Once a firm’s reputation is tainted, people find out it’s too hard to do business or too hard to get through the application… You can’t even really calculate that.” Fitzgerald explained.

In the era of AI-enabled fraud, effective compliance is not only about stopping bad actors. It is about improving fraud operational efficiency, protecting reputation and unlocking revenue growth.

Why Future Proof AI Fraud Prevention Means Applying Intelligent Friction 

Both Cohen and Fitzgerald emphasized intelligent verification as the foundation of modern AI fraud prevention.

“You want to make sure that you’re not slamming everyone with the same kind of verification flow,” Cohen said. “You want to challenge certain high-risk users in different ways and allow low-risk users to come through faster.”

Intelligent friction applies verification dynamically based on behavior, context and regional risk patterns. Risk landscapes differ by geography, demographic and transaction type. A globally consistent strategy, tuned locally, improves fraud operational efficiency while reducing unnecessary onboarding friction. From synthetic identities and deepfakes to automated agent-driven attacks, this adaptive approach helps organizations address the full spectrum of modern threats.

In the era of AI-enabled fraud, smarter identity is not simply defensive. It is strategic.

How to Implement Modern Fraud Prevention Solutions

Modern AI fraud prevention operates in real time. In live payments environments, every transaction should be evaluated instantly to separate malicious automation from the legitimate automated activity customers rely on.

For companies like WEX, this means evolving fraud strategies beyond static rules and manual review. Effective AI fraud prevention now relies on layered signals, behavioral intelligence and adaptive decisioning to detect automated fraud without slowing good users.

Identity frameworks are evolving as well. Alongside Know Your Customer (KYC) and Know Your Business (KYB), Know Your Agent (KYA) is emerging as a critical third pillar for the agentic era. As AI agents begin executing transactions on behalf of users, KYA verifies who built the agent, what code is running and whether the user authorized its actions.

“It’s a really exciting time to be in this space,” Cohen said. “We’re seeing protocols emerge that enable transactions between agents, merchants and payments systems.”

What Digital Trust Looks Like in the Age of AI Fraud 

Today’s threat landscape is defined by AI-enabled fraud operating at scale, and legacy verification systems increasingly expose organizations to significant revenue and reputational risk.

Addressing this challenge requires more than new tools – it demands a shift in mindset. Fraud prevention is no longer a siloed responsibility for a single team. In an era of automated fraud, protecting digital ecosystems requires coordination across product, risk, compliance and growth. 

Modern identity strategies built on AI fraud prevention, layered signals and intelligent friction help organizations reduce fraud losses, minimize false positives and improve fraud operational efficiency across the business. The impact extends far beyond onboarding: stronger identity infrastructure simplifies downstream operations, reduces manual investigations and enables teams to focus on growth.

As AI-driven commerce accelerates, organizations that embed fraud prevention into their broader operating model will be best positioned to scale trust, protect revenue and compete in an increasingly automated digital economy.