iDenfy, a prominent RegTech firm known for its expertise in identity verification and fraud prevention, has unveiled enhancements to its Know Your Customer (KYC) and Anti-Money Laundering (AML) platform. These updates address the growing issue of synthetic identity fraud, an increasing problem that has reportedly surged by at least 15% across various sectors, notably impacting high-risk areas like iGaming.
Amidst rising fraud rates, iDenfy emphasises the importance of integrating human expertise with technology. While software solutions are critical, they must be supplemented with manual reviews to effectively combat synthetic ID fraud. By sharing insights from its recent software updates, iDenfy spotlights its dual approach.
Role of skilled personnel
Domantas Ciulde, CEO of iDenfy, highlighted the company's substantial investment in its team, particularly during the first quarter of 2026. He noted that the expertise of the in-house KYC team is crucial for refining the product and ensuring thorough manual reviews of ID document data. This method mitigates false positives, enabling swift service access for genuine users. iDenfy's KYC specialists operate around the clock, ensuring continuous service without downtime.
Deepfake videos are used as disguises during biometric verification, further complicating fraud detection
Detecting synthetic identity fraud is becoming increasingly challenging. Fraudsters use real data, like Social Security numbers, coupled with fictitious names and AI-generated images to bypass simple identity verification systems. Additionally, deepfake videos are used as disguises during biometric verification, further complicating fraud detection.
Automation versus manual review
In response to these sophisticated attacks, traditional rule-based and AI-only security systems focus on known threats but struggle with new patterns generated by advanced AI fraud tools. However, fully relying on human review is not feasible due to the high volume demands of digital onboarding. iDenfy balances this by employing a hybrid approach.
iDenfy's system, upon receiving identity documents, conducts an immediate check across a vast database of over 3,000 government-issued document types covering 200+ regions. Preliminary checks, including facial recognition and liveness detection, are completed within three minutes. These steps are crucial in filtering out deepfakes before moving forward in the process.
Manual verification processes
iDenfy extends its services beyond initial verification through ongoing AML screening
Unlike platforms that rely on human review as a secondary measure, iDenfy's compliance team manually assesses each identity audit, going beyond simply addressing flagged issues. The team is trained to identify unique fraud patterns and anomalies that automation may miss, ensuring robust security measures.
iDenfy extends its services beyond initial verification through ongoing AML screening. This process involves constant checks against global sanctions, Politically Exposed Persons (PEP) lists, and adverse media to maintain compliance and mitigate reputational risks. For business clients, iDenfy's AI Company Reviewer automates KYB decisions, flagging suspicious cases for manual inspection.
Challenges of synthetic identity fraud
The economic impact of synthetic identity fraud is significant. In 2025, US lenders suffered losses exceeding $35 billion due to fake identities. Globally, synthetic identity fraud poses a threat of potential losses ranging from $20 to $40 billion annually, exacerbated by the capabilities of generative AI tools.
In summarising iDenfy's strategy, Domantas Ciulde stated, “Manual company reviews have been one of the most resource-intensive parts of compliance, and fully automated systems create a different problem: they can’t adapt in real-time to fraud patterns they’ve never seen. Our model is built on the premise that AI and human expertise are not interchangeable. They’re complementary. AI handles speed and scale. Our team handles difficult scenarios.”
iDenfy, a global RegTech specialising in identity verification and fraud prevention solutions, shares its newest insights into the synthetic identity fraud crisis, introducing vital adjustments to its Know Your Customer (KYC) and Anti-Money Laundering (AML) platform. Company officials claim that fraud rates have increased at least 15% in various industries, especially in high-risk sectors like iGaming.
As a result, the best way to actually combat issues like synthetic ID fraud or gen AI fraud is not to entirely rely on software but to mix it with human reviews. For that, iDenfy provides its services as an example, allowing readers to take a peek at the newest software adjustments.
AI system errors
According to Domantas Ciulde, the CEO of iDenfy, the company has been investing in its team, the best asset behind the identity verification product, the most in the first quarter of 2026, stressing that skilled personnel are the key to not missing details when it comes to product development. A huge part of valuable insights and suggestions comes from the in-house KYC expert team, who are responsible for manually reviewing captured ID document data.
False positives can affect genuine users trying to access services quickly. A manual approach helps correct AI system errors and onboard legitimate users almost instantly. For example, iDenfy’s KYC specialists work 24/7, removing the risk of downtime or long queues. While this manual approach to KYC in a world of AI seems like a nice-to-have factor for some ID verification service providers, Domantas Ciulde disagrees:
Customer onboarding process
“A highly skilled in-house team is extremely valuable. We started with a few members manually re-checking documents and built a full team from scratch. This means no third parties, just training on real forged document cases that we constantly detect when servicing our clients.”
Regarding synthetic identity fraud in particular, it becomes a more complex task to detect bypassing attempts during the customer onboarding process. Sophisticated fraudsters outsmart basic IDV systems by combining real data, such as valid Social Security numbers (SSNs) or real addresses, with fake names, borrowed birth dates, or AI-generated document images. Deepfake videos are also becoming more advanced, with fraudsters using them as a virtual mask in the second KYC layer, biometric verification, where selfie data needs to be cross-matched with the photo on the submitted ID document.
Modern digital onboarding demands
In this sense, the verification industry’s default response with more automation has not closed the gap. Reported identity and fraud losses in the US from the prior year in automated detection grew.
The structural problem is that traditional, rule-based, and AI-only systems defend against known attack patterns. On the other side, generative AI produces unknown ones, purpose-built to defeat the specific model being targeted. At the same time, only manually performed review processes are economically unscalable at the volume that modern digital onboarding demands.
AI-generated biometric media
In general, when a new user submits their identity documents, iDenfy’s solution immediately cross-references against a database of more than 3000+ government-issued document types from over 200+ countries and territories. Initial recognition and 3D liveness detection check if the biometrics of the face match in under 3 minutes. Critically, iDenfy’s KYC system filters deepfakes at the beginning point of user registration and catches AI-generated biometric media before it moves on to the next step.
This layer handles the speed and scale challenge that no manual process can match. Usually, the first layer is enough for most of the new user onboarding situations, except for the cases when the documents are not clear or look suspicious.
Suspicious contextual anomalies
Unlike platforms that treat human review as a fallback, iDenfy’s in-house compliance team then manually reviews every identity audit, not only the flagged ones.
Reviews are completed by double-checking the similarities between the document and the submitted face, as well as specific countries where the user can be marked as a red flag. iDenfy’s model diverges most sharply from automation-only competitors without a real regulatory background due to the trained and experienced compliance review team that can catch the specific cases, new fraud patterns, and suspicious contextual anomalies that no AI model can bypass.
Meeting compliance requirements
The third layer is AML screening and ongoing monitoring. Identity verification is not only a one-time security guarantee after onboarding the user. iDenfy’s platform offers continuous screening solutions against global sanctions lists, Politically Exposed Persons (PEP) databases, and adverse media sources throughout the user’s active account journey.
This is done to ensure that the person can meet compliance requirements, help to identify suspicious patterns that require further analysis, and save from massive reputational damage. For regulated business clients, iDenfy also launched its AI Company reviewer, which automates Know Your Business (KYB) decisions by cross-referencing company documents and completing other important tasks like Ultimate Beneficial Owner (UBO) verification using government registry data. Suspicious cases are flagged in seconds. Cases with partial matches or flagged signals are immediately routed to the human review team to ensure that no further mistakes are made from the company’s side.
Generative AI fabrication tools
It’s worth mentioning that in 2025, lenders in the US lost more than $35 billion because of fake identities that were linked to new bank accounts. This is a big problem for the whole world. It’s thought that businesses could potentially lose between $20 and $40 billion every year to synthetic identity fraud attacks that are happening when fraudsters use powerful generative AI fabrication tools.
Synthetic identity fraud typically begins with identity fabrication, in which scammers create a hybrid persona based on a stolen SSN that lacks an active credit file and a fake biometric identity in order to pass verification systems and make it more difficult to detect the account.
“Manual company reviews have been one of the most resource-intensive parts of compliance, and fully automated systems create a different problem: they can’t adapt in real-time to fraud patterns they’ve never seen. Our model is built on the premise that AI and human expertise are not interchangeable. They’re complementary. AI handles speed and scale. Our team handles difficult scenarios,” added Domantas Ciulde, the CEO of iDenfy.
