Identity Fraud Intelligence: Trends & Insights, pairs network-scale fraud data with market intelligence to show where identity fraud is heading and where each market sits on the global curve.
NEW YORK, July 15, 2026. Liminal, the actionable intelligence company, and Unico, a global identity network, this week released Identity Fraud Intelligence: Trends & Insights, a co-authored report on how identity fraud is changing and what it now takes to defend against it. Its central finding is an economic one: the cost of running a sophisticated attack has fallen more than 100x, while the cost of defending against it keeps climbing. Liminal projects fraud losses at financial institutions will rise 121% by 2030, from roughly $25 billion to $55.3 billion.
The scale is already substantial. Global fraud losses now exceed $400 billion a year by industry estimates, with generative AI accelerating the climb. What was once a line item institutions could reserve against has become a structural cost of doing business.
The harder problem is that defenses are organized the wrong way for it. Most institutions still run identity verification and fraud detection as separate systems, even as attackers move fluidly across the identity lifecycle and from one institution to the next. The same actor rejected at one institution simply arrives at the next as a first-time customer. Liminal’s research finds that 93% of practitioners express confidence in their fraud models, yet 92% acknowledge that legacy infrastructure still lets fraud signals slip through. That gap, between confidence and connected defense, is what the next phase of fraud will exploit.
“Fraud now operates as an industry, with shared tooling and a market of its own, not the predictable cost institutions used to budget for,” said Filip Verley, Chief Innovation Officer at Liminal. “The institutions that pull ahead over the next 18 months will be the ones that connect identity and fraud into a single system, because the attackers already operate that way. This report gives leaders a clear read on where their own market sits on that curve.”
“Seeing fraud across an entire network, rather than one institution at a time, changes the picture completely,” said Davi de Castro Reis, VP of Global Strategy at Unico. “To permanently stop serial fraud, we must associate identity with network-based biometric verification backed by real-time liveness. It is the only way to expose the underlying fraudster moving across institutions. In our data, a single fraud entity was tied to 949 different identities and reached as many as 30 businesses, appearing as a stranger to each one.”
The report’s findings include:
- The sophisticated tier of attacks, combining deepfakes, injection, and physical manipulation, now makes up 23.3% of classified attacks, up sharply from a year earlier.
- 85% of organizations report repeat account-takeover incidents, nearly half of them frequently, and 62% expect their compliance budgets to grow.
- 76% of practitioners now rank device fingerprinting as the most effective defense against AI-enabled fraud, and strengthening partnerships is the fastest-rising 2026 investment priority at 47%.
The report also introduces a five-rung maturity model that lets institutions place their own defenses and identify the gap that most predicts their exposure over the next 18 months.
Identity Fraud Intelligence: Trends & Insights was published on June 25 and is available to download here. Liminal and Unico will host a joint webinar on the findings on July 29. Registrations are open and free.
