AI is changing the economics of deception. It is now cheaper, faster, and easier to manufacture credibility — not just stolen credentials, but entire identities that look real enough to pass onboarding systems, financial controls, vendor reviews, remote hiring processes, and other trust-based workflows.
That is what makes synthetic identity fraud so dangerous. It is not only a financial crime issue, and it is not only a cybersecurity issue. It is a risk intelligence challenge with cybersecurity, operational, reputational, and financial consequences. When an adversary can manufacture a believable identity, they can create the appearance of trust before an organization recognizes the underlying risk.
Synthetic identity scams have recently been recognized as the fastest-growing type of fraud globally. According to the LexisNexis Risk Solutions 2026 Cybercrime Report, cyber-based fraud rose at a staggering rate of 8% in 2025, more than one in ten cases of which involved the use of synthetic identities — an eightfold increase year-over-year and clear indication that conventional cybersecurity strategies are simply insufficient in the age of AI.
What is Synthetic Identity Fraud?
For decades, consumers have lived under the threat that their identities could be stolen and used to drain their own bank accounts or tank their credit scores. While this threat certainly hasn’t gone away, stealing someone’s actual identity is no longer necessary for cybercriminals to successfully perpetrate fraud and on a larger scale than ever before.
Thanks largely to rapid advancements in AI over the past few years, it’s become increasingly easy for cybercriminals to simply fabricate and nefariously leverage fictional identities. Fraud is ultimately an economic activity: criminals weigh cost, complexity, risk, and potential return. By significantly lowering the cost to create believable identities — complete with resumes, documents, images, video, profile histories, and interaction patterns — generative and agentic AI make the return on fraud more attractive. The larger danger is not simply that AI can create fake people. It is that AI can manufacture enough supporting context to create synthetic trust. When the cost of deception falls this dramatically, the volume of fraud does not grow incrementally. It proliferates. Once this occurs, synthetic identities can be used to infiltrate organizations via remote job opportunities or gain access to financial accounts and/or sensitive data and corporate infrastructure.
Assembled from breached Social Security numbers and public records, these synthetic identities are typically carefully curated over months before being used to build fabricated histories and credible financial footprints that allow cybercriminals to easily slip through the cracks of traditional identity verification processes.
This is what makes it such a significant challenge for organizations: as opposed to conventional identity fraud, there isn’t any real “person” to report themselves as a victim of a crime. As a result, these identities can persist undetected for months, while being used to extract larger sums of money and exposing organizations to steadily compounding risk.
Increasing accessibility to sophisticated, public-facing AI tools is central to this new mode of deception; personalized AI agents are, by definition, programmed to the user’s specifications. This includes generating highly realistic yet entirely fabricated identities, oftentimes complete with deep-faked videos and other content submitted to an organization or posted on social media, then further utilized to automate financial fraud or cyberattacks.
Once almost exclusively associated with North Korean hackers, these tactics are now being leveraged on a much broader scale and by non-state actors to exploit companies across industries.
The Best Offense is a Strong Defense
Organizations must move now to counter synthetic identity fraud as it’s a very real, increasingly serious, and pervasive threat. The average valuation of corporate fraud via data breaches now sits at roughly $4.4 million, and regulatory agencies such as the SEC are already tightening fraud report deadlines and applying more scrutiny on factors including cybersecurity and third-party risk management.
Quite simply, periodic financial reviews are no longer a sufficient risk management strategy in today’s rapidly evolving technological ecosystem.
Without proactive, real-time threat monitoring, organizations risk:
- Manufactured identities entering customer, employee, contractor, vendor, or partner workflows
- Delayed detection of active campaigns, allowing attacks to mature before they are identified
- Blind spots beyond the network perimeter where adversaries plan and coordinate attacks
- Low-quality threat hunting hypotheses driven by assumptions rather than current threat intelligence
- Wasted time and effort chasing benign anomalies without contextual validation
- Increased operational and financial impact as threats are discovered only after damage occurs
A good offense via AI-enabled risk intelligence is key to a successful defense against modern cybercrime threat vectors. In the same way organizations are actively embedding intelligence into daily operational workflows to improve decision-making and productivity, AI-based technologies also enable real-time monitoring and improved detection of suspicious activity such as synthetic identity fraud, which can often evade traditional cybersecurity defenses.
Open-source intelligence and commercially available information can play a critical role in detecting synthetic identities because they help organizations understand whether an identity makes sense in context. Detecting synthetic identity fraud requires more than checking whether a field matches a database. It requires asking whether the broader pattern holds together. Does the digital footprint have depth and age? Do relationships, locations, affiliations, and behaviors align? Are there inconsistencies across public, commercial, and behavioral signals? Does the identity connect to other entities in ways that are credible, explainable, and consistent over time?
AI-enabled risk intelligence can help organizations fuse these signals at scale. It can support entity resolution across fragmented data, surface inconsistencies, identify anomalous patterns, score confidence, and package evidence for human review. The goal is not to replace human judgment. It is to give investigators, analysts, fraud teams, security teams, and risk leaders a faster, richer, and more defensible way to understand identity risk.
This is where the response to synthetic identity fraud should evolve. Traditional cybersecurity controls remain necessary, but they are only part of the answer. Synthetic identities attack the assumptions underneath those controls: that the person, vendor, applicant, account, or digital persona being evaluated is who they claim to be.
As cybercriminals and fraud networks become more sophisticated, organizations need to move from reactive detection to proactive discovery. In an era when AI can manufacture trust, static checks and after-the-fact reviews are not enough. Organizations need continuous, evidence-backed intelligence that can connect signals, expose inconsistencies, and help humans make faster, more confident decisions.
The future of fraud prevention will not be won by asking whether an identity looks real at a single moment in time. It will be won by understanding whether that identity behaves, connects, and evolves like a real person — and by detecting manufactured risk before it becomes organizational exposure.
_____
About the Author
John Larson is President and Chief AI Officer at Babel Street. A nationally recognized AI and technology leader, Larson has more than two decades of experience building high-performance AI organizations. He previously was Executive Vice President and Head of Artificial Intelligence at Booz Allen Hamilton, where he played a pivotal role in establishing and reinforcing the firm’s position as the leading provider of AI services to the U.S. federal government. His impact earned him recognition as one of the Top AI Executives to Watch in 2025. While at Booz Allen Hamilton, he launched a series of reusable mission-aligned AI products, established industry-shaping partnerships with NVIDIA, AWS, and next-generation AI start-ups, and built responsible AI governance frameworks that accelerated secure, explainable AI adoption across defense, intelligence, and civil agencies. Prior to this, Larson held senior leadership roles at IHS Markit (now S&P Global), Deloitte, IBM, and PwC.
Click Here For The Original Source.
