Directory of AI scam and authorised push payment fraud vendors

Charm Security sells scam and social engineering defence to banks, fintechs, payment providers and credit unions. Rather than scoring transactions or devices, the platform assesses what the company calls human vulnerability exposure, analysing customer risk patterns using psychological insight to identify who is susceptible to manipulation, then deploys agents that intervene in real time while a scam is unfolding across digital, voice and in person channels. It was created inside Team8's venture creation fund and its stated regulatory driver is the shift of authorised push payment fraud liability onto banks.

Cybera sells banks, crypto platforms and cyber insurers two products aimed at authorised payment scams. Mule Intelligence feeds verified accounts and wallet addresses used by scammers into an institution's existing fraud and compliance systems so outgoing payments can be blocked in real time and mule accounts inside its own customer base surfaced, with the intelligence gathered from active defence operations, victim reports and law enforcement collaboration rather than inferred from behaviour, which the company calls non probabilistic. Scam Response handles the aftermath: a victim reports online and within minutes the case is dispatched to law enforcement, beneficiary banks and exchanges, with a dedicated team managing victim communications on the institution's behalf. The company claims recovery chances improve up to tenfold.

Early Warning Services is a bank owned financial infrastructure company based in Scottsdale, Arizona, wholly owned by seven of the largest United States banks and operating for more than three decades under the name Primary Payment Systems before rebranding. It runs three brands: the Zelle payments network, the Paze digital wallet, and Certos, the unified portfolio of identity, account and payment risk products launched in April 2026 that carries the inference line and is the basis on which this record is graded. Certos sells to roughly 2,500 banks and credit unions and to about 5,000 total participants spanning financial services companies, payment processing companies, merchants, government entities and identity theft protection service providers. The products are Verify Identity, which scores in real time the likelihood that an applicant is who they claim to be; Predict New Account Risk, which draws on contributed deposit account activity to surface early indicators of account misuse and first party fraud; Deposit Chek, which evaluates deposit risk for funds availability decisions using embedded machine learning models that assess the likelihood of check and automated clearing house returns within thirty days; Payment Chek, which screens payments and disbursements with real time proprietary risk scoring before funds are released; Verify Account, which confirms account ownership and status for funding, transfers and disbursements; Expand Credit Insights, which adds deposit account context to lending decisions for applicants with no traditional credit history; and Asset Search and Verification, which supports government eligibility determinations. All of it runs over the National Shared Database, a consortium resource contributed to by participating institutions under a give to get model, for which the company holds the role it calls Trusted Custodian. In 2025 the portfolio screened more than 11.4 trillion dollars in payment and deposit transactions, supported 124 million new account applications and reports preventing more than 3 billion dollars in potential fraud loss. Early Warning is separately a nationwide specialty consumer reporting agency under the Fair Credit Reporting Act, listed on the federal consumer bureau register of consumer reporting companies, which is the source of its unusual privacy, regulatory and recourse posture. In May 2026 it partnered with an indexed identity decisioning platform to distribute the Certos suite into community banks and credit unions.

FiVerity runs an anti-fraud collaboration platform letting banks, credit unions, payment institutions and online lenders share fraud intelligence with each other, with regulators and with law enforcement, on the premise the Federal Reserve itself has stated: no single organisation can stop synthetic identity fraud alone. Its distinguishing design is how the sharing works. Institutions exchange pattern matches rather than consumer personal information, protected by double blind encryption, so fraud intelligence moves without customer identities moving with it. The platform aggregates alerts across the network, enables joint investigations under the statutory information sharing safe harbour, pre-fills suspicious activity reports, and surfaces repeat offenders operating across multiple institutions. It deploys without integration work, and claims to catch more than half the fraud conventional rules based systems miss.

Heka assembles intelligence about individual consumers from outside an institution's own records, drawing on live open web data, digital footprint analysis, darknet sources and non reporting collections data across thousands of global sources, and structuring it into profiles that surface alias use, reputational exposure and behavioural anomalies. Banks, insurers, payment processors and pension schemes use those signals for fraud detection, credit and insurance underwriting, onboarding and consumer tracing, delivered through a single interface or in batch and returned inside 300 milliseconds for transaction decisions. The company describes its approach as drawn from intelligence community tradecraft and positions explainability and auditability as central, on the argument that credit bureau files and velocity models miss what is happening online.