Artificial intelligence may be boosting productivity everywhere from banks to government agencies, but it’s also rewriting the rulebook for deception. What used to take a skilled fraudster hours—or days—now happens in seconds with a halfway-competent model and a laptop. Just in time for International Fraud Awareness Week 2025, the Association of Certified Fraud Examiners (ACFE) and SAS are sounding the alarm.
In a new cross-industry survey previewing the fourth edition of the Anti-Fraud Technology Benchmarking Report, the findings are stark:
- 77% of anti-fraud professionals report a spike in deepfake-driven social engineering over the last two years.
- 83% expect those attacks to increase further by 2027, with over half predicting a significant jump.
If that sounds like a bad plot twist in a cyber-thriller, it’s not. It’s an emerging consensus in the fraud-fighting community: the next era of financial crime won’t just use AI—it will depend on it.
The Deepfake Era Arrives (Ready or Not)
“Artificial intelligence has become one of the most powerful tools in business—and one of its most potent threats,” said ACFE President John Gill. It’s a sentiment increasingly echoed across compliance teams, banks, governments, and insurers grappling with attacks that can imitate voices, clone faces, and forge documents with uncanny accuracy.
The survey also surfaces another troubling reality: fewer than 1 in 10 anti-fraud professionals feel fully prepared for AI-accelerated threats. That readiness gap is quickly becoming one of the defining cybersecurity challenges of the decade.
To help close that gap, ACFE and SAS are offering a Fraud Week webinar—Agentic AI in Action: Intelligent, Adaptive Fraud and Financial Crime Prevention—on Nov. 17 at 11 a.m. ET. It’s designed to equip practitioners with the latest tactics before AI-generated fraud hits its next growth spurt.
Why AI Fraud Is Exploding—and Why It Matters
AI-powered fraud is evolving for the same reason AI-powered productivity is booming: access.
Large models are cheaper to run, easier to fine-tune, and increasingly capable of producing realistic synthetic media. Combine that with a global shift to digital identity, remote transactions, hyperconnected supply chains, and real-time payments—and fraudsters suddenly have a playground rich with opportunity.
Stu Bradley, SVP of Risk, Fraud and Compliance Solutions at SAS, frames the challenge this way:
“AI is blurring the boundary between truth and imitation, with untold billions at stake.”
It’s a problem that spans industries:
- Financial institutions are fighting hyper-realistic deepfake attacks on onboarding and KYC systems.
- Governments are battling identity-theft-driven fraud in public benefit programs.
- Insurers are dealing with AI-generated documents, synthetic identities, and increasingly sophisticated criminal networks.
- Retailers and e-commerce platforms are absorbing a wave of automated refund fraud and bot-amplified scams.
In short: if your organization authenticates people, processes claims, or moves money—you’re in the blast radius.
Inside the Industry Response: Field-Proven Approaches to AI-Age Fraud
While fraudsters have found new superpowers, industry players haven’t been idle. SAS, long a top provider in risk and fraud analytics, is spotlighting how organizations worldwide are deploying machine learning, network analytics, and real-time scoring to stay ahead.
Below, we break down several real-world examples—ranging from national digital ID systems to UAE banks and Korean insurers—that highlight how institutions are adapting to the new threat landscape.
1. Banking: AI as Both Adversary and Hero
Banks are among the most targeted sectors—and among the fastest adopters of AI defense tools. SAS points to one global bank that cut alerts by 40% while improving detection accuracy by 35% through advanced analytics.
That double win—reduced noise, improved signal—is becoming the benchmark. The industry is moving from after-the-fact fraud detection to anticipatory defense, powered by behavior-based models that learn and adapt in real time.
2. Public Sector: Protecting Massive, High-Stakes Programs
Government programs remain prime targets for identity theft and synthetic fraud, especially in high-volume benefits programs.
SAS’ Public Sector Fraud-Fighting Maturity Assessment helps agencies benchmark where they stand and identify readiness gaps. And with generative AI making fraudulent documentation easier to produce, public programs are accelerating adoption of risk-scoring and machine learning tools.
When billions in taxpayer dollars are at stake, “good enough” is no longer good enough.
3. High-Trust Digital Identity: Norway’s BankID Moves to Real-Time Defense
If you want a glimpse of the future of national-scale digital identity, look north.
BankID, Norway’s national digital ID provider, processes nearly a billion identity transactions each year—touching everything from banks to government portals to private-sector logins.
Now BankID is integrating its identity intelligence directly into SAS’ real-time fraud scoring engine, turning decades of verified identity data into instant behavioral analytics.
The result?
- Earlier anomaly detection
- Better risk scoring
- Stronger defenses against account takeover and synthetic identity fraud
As BankID’s David Sæle explains, “We’ve moved from reacting to fraud to anticipating it.” That shift—from defensive to predictive—may soon define digital identity worldwide.
4. Real-Time Transaction Monitoring in the UAE
Ajman Bank, a fast-growing Sharia-compliant bank in the UAE, is also embracing AI-enabled fraud prevention. With SAS and regional partner DataScience Middle East, the bank deployed real-time monitoring across cards, payments, and digital channels.
Machine learning models now score behavior as it happens, helping the bank:
- Reduce false positives
- Prioritize high-risk threats
- Consolidate analytics across channels for a unified fraud defense
For a region experiencing explosive digital payments growth, the timing couldn’t be better.
5. Korea’s First AI-Powered Insurance Fraud Network
Insurance fraud isn’t new—but detecting organized criminal networks at scale is.
DB Insurance, covering more than 10 million customers, built Korea’s first AI-powered fraud network using SAS Viya. The system connects decades of claims, policy, and customer data, then applies network analytics to uncover hidden relationships between bad actors.
The impact is dramatic:
- Detection accuracy jumped 99%
- Case analysis time fell from hours to minutes
- Case throughput increased 30x
One investigator described it succinctly: “Dozens of invisible fraud connections lit up instantly.” It’s a first-of-its-kind model likely to be replicated across the industry.
6. Strengthening SNAP Payment Integrity in the U.S.
On the government side, one major U.S. state has spent years modernizing its SNAP (food assistance) oversight with SAS:
- Workflow automation reduced backlogs
- Machine learning models risk-scored overpayment referrals
- The system now flags high-risk cases across all active recipients
The result:
- 50% reduction in investigation processing times
- Processing windows cut from 12 months to 6
- More precise reviews despite tight budgets and staff limits
It’s a compelling blueprint for modernizing oversight in public assistance programs nationwide.
The Big Picture: The Future of Fraud Is AI-Driven—and So Is the Defense
The new ACFE–SAS research paints a clear picture: the fraud landscape is shifting faster than most institutions can keep up. Deepfakes, synthetic identities, and automated scams are no longer fringe threats—they’re mainstream tools.
But the industry isn’t powerless. High-trust digital identity signals, real-time risk scoring, network-level analytics, and agentic AI models are giving defenders new ammunition.
Still, the preparedness gap remains. With only a small fraction of professionals feeling ready, the next two years may define whether industries can adapt quickly enough to outpace AI-driven fraud—or whether attackers gain a lasting advantage.
For now, the message is clear: awareness is essential, collaboration is critical, and AI is both the problem and the solution.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
