February 2025 Summaries

Socure argues that reliable address verification is essential to Customer Identification Program and anti-money-laundering compliance because PO boxes and commercial mail receiving agencies can prevent financial institutions from locating customers. Its testing found that even USPS tools can omit or misclassify postal facilities, illustrated by a New Mexico post office that appeared as an ordinary street address and was not returned by USPS API or nearby-facility searches. In response, Socure developed a method that cross-references multiple data sources, iterates through postal codes and location identifiers, evaluates carriers such as FedEx, UPS, and DHL, and uses its identity graph to correct commercial classifications caused by missing unit numbers. The company says this approach reduced false positives for PO box and CMRA flagging from 0.5% to zero, reinforcing its view that verification systems must account for the differing strengths and limitations of ostensibly authoritative data sources.

Socure presents Verify as a non-documentary identity verification platform designed to address shortcomings of legacy credit-data-based systems, particularly for Gen Z, new-to-country consumers, rural residents, and others with limited traditional credit histories. It combines authoritative and alternative sources such as education, utilities, tax, telecommunications, payment, and alternative credit records with machine learning, entity resolution, natural language processing, and extensive name-matching capabilities to create a persistent, multidimensional consumer profile. Socure claims verification rates of up to 99% for mainstream populations and 96.3% for Gen Z consumers, alongside expanded detection of deceased identities and coverage of more than 102 million non-USPS addresses. The company says its approach reduces false rejections and manual reviews while providing risk codes, matched data, normalized addresses, and other metadata for compliance and automated decision-making. Verify also supports specialized use cases such as identifying sports-league personnel for betting operators and has expanded passive identity verification coverage to more than 50 countries, with options to combine its service with document verification.

Socure presents Sigma Synthetic as an identity fraud detection model designed to identify manipulated and fabricated synthetic identities during onboarding, high-risk authentication, and portfolio screening. The company reports that its latest model captures 74% of fraud at a 3% risk depth and 59% at a 1% risk depth, while claiming substantially greater accuracy than competing synthetic fraud and credit bureau tools. Sigma Synthetic combines historical and third-party data, expert human labeling, Social Security number metadata, and “proof of life” records such as property, driver’s license, and education data to detect inconsistencies and verify thin-file applicants. Socure says the approach can reduce fraud while limiting friction for legitimate users, including younger and new-to-country consumers, and that regular portfolio screening can uncover dormant threats and compliance risks. The enhancements are available to existing customers without added cost, and a cited retail-bank study estimated potential annual savings of about $27 million in fraud losses and $20 million in added revenue from fewer false positives.

Socure describes updates to its Sigma Identity Fraud platform, which uses a network identity graph drawing on data from more than 2,800 organizations to assess identities through historical behavior, device and session signals, and personally identifiable information rather than relying on static checks. Its Entity Profiler and Integrated Anomaly Detection features are intended to identify patterns associated with identity theft, device farms, email manipulation, muling, VoIP use, and AI-enabled fraud, including at onboarding and high-risk authentication. The company reports that Sigma captures up to 85% of third-party identity fraud within the riskiest 3% of applications and 89% within the riskiest 5%, compared with a stated industry competitor average of 37% for the former group. Socure also cites a credit card issuer analysis estimating that its system could have increased fraud capture from 57% to 82%, saved more than $4.7 million annually, reduced false positives, and auto-approved 98% of users previously subject to friction. The company says model improvements are provided to customers without additional charges and emphasizes ongoing investment as fraud tactics evolve.

Romance scams caused more than $1 billion in reported U.S. losses in 2023 and typically involve criminals using fabricated online identities to build emotional trust before requesting hard-to-recover payments through wire transfers, gift cards, peer-to-peer apps, or cryptocurrency. Scammers often use love bombing, overseas or military cover stories, invented emergencies, and promises of future wealth to discourage verification and prompt urgent financial assistance, while generative AI and deepfakes have made fake profiles and video interactions more convincing. People who are isolated, newly relocated, widowed, divorced, or otherwise emotionally vulnerable may face increased risk, and victims can experience severe financial loss, shame, social withdrawal, or even unwitting involvement in money laundering. Warning signs include refusal to meet or video chat, inconsistent personal details, rapid declarations of love, requests for secrecy or money, and profiles whose images or claimed occupations cannot be independently verified. The text argues that dating and social platforms should supplement user education with integrated identity verification, behavioral analysis, device intelligence, biometric liveness checks, and shared fraud data to identify manipulated or repeat identities before they can target users.

Socure and Effectiv have introduced RiskOS®, a unified platform for fraud prevention, identity verification, and regulatory compliance following their combination in 2024. The platform combines Socure’s identity graph with Effectiv’s decisioning engine, integrating more than 20 Socure solutions and 50 partner products to replace fragmented risk-management tools across the customer lifecycle. Initial applications include consumer and business onboarding, bank-account verification, account-takeover prevention, and trust-and-safety assessments, with workflows designed to support real-time, stateful risk decisions. RiskOS® offers a no-code drag-and-drop interface, prebuilt templates, customizable workflows, A/B testing, and integrations for enterprise machine-learning models, while claiming execution times below 150 milliseconds and throughput above 1,000 queries per second. Planned 2025 additions include anti-money-laundering transaction monitoring, payment-fraud detection, and credit underwriting, with the stated goal of improving fraud detection while reducing friction for legitimate customers.

Socure reports that its DocV document-verification platform improved speed, fraud detection, accuracy, accessibility, and user completion rates during 2024 through features including deepfake selfie detection, injection-attack defenses, barcode reconstruction, improved OCR, and a server-driven v5 SDK supporting selfie reverification and secondary documents. The company says its barcode-reading performance rose 65%, unreadable or unclassified IDs fell 50% year over year, language support expanded to nearly 30 languages, and latency declined 91% from 15 seconds in 2023 to a P95 of 1.26 seconds. Reported 2024 results include a 98.2% true-accept rate, 98.8% true-reject rate, 1.2% false-accept rate, 1.8% false-reject rate, 90.4% overall approval rate, and 3.2% resubmission rate, meaning most users received a conclusive result on their first attempt. Looking ahead, Socure plans to expand barcode-only verification and strengthen defenses against evolving deepfakes, while positioning the platform as a way for customers to reduce fraud, manual reviews, operational costs, and onboarding friction.