Cybera

Every listed account is verified through direct interaction and investigation, so the defining asset is human intelligence gathering rather than a model. Models handle throughput, processing scam reports and surfacing patterns across mule networks at a volume no team could read.

Matches inform decisions the bank already owns, and case dispatch to authorities is automated while people handle victim communications. How an account is reviewed before it enters the feed is not described.

Proven accuracy and zero guesswork are claimed, yet no accuracy figure, verification procedure or evidence standard is published. For an output that brands accounts as criminal, how often a listing proves wrong is the missing disclosure.

More than 50 million customers secured, $10 million recovered and 10,000 victims assisted are real aggregate figures. Banks, exchanges and insurers are never named as customers, and no figure is joined to an institution.

Scams suffered by one institution's customers produce watchlist entries every other subscriber uses, which is the mechanism's value. What an institution contributes through the response service, what consent victims give and how long entries persist are not stated.

Victim reports carry the circumstances of a loss with personal and account details, and the feed names specific accounts and wallets as belonging to scammers across many institutions. The privacy policy, updated May 13, 2025, covers data controlled through the website and app and expressly excludes service provider processing.

Independent auditors are said to assess security, with no standard, auditor or report named. Certifications and a trust portal are absent from the homepage, product and company pages.

Anti money laundering pressure is referenced in general terms, with no supervisor, statute or instrument named. Advisors Marcel Rohner, president of the Swiss Bankers Association, and Craig Jones, formerly of INTERPOL, bring experience rather than a regulatory position, and mandatory push payment reimbursement is not mentioned.

Identifying mules by investigation rather than behavioral scoring avoids freezing people whose only offense is transacting unusually. Mule account holders are often coerced or recruited victims themselves, and listing accuracy is not evaluated in anything published.

Wrongly listed account holders can have payments blocked or accounts closed across every subscribing institution, yet no evidence behind a listing, review or removal process, or route to contest is described. Scam Response helps victims recover money, which is the product's function rather than recourse for an error in the feed.

Active defense operations, victim reports and enforcement partners are the sources, described by type only. Partners, data suppliers and model providers are unnamed, and no subprocessor list is published.

An API delivers matches into transaction monitoring, risk scoring and suspicious activity reporting, but no platform, vendor or system is named. Developer documentation is not linked.

Switzerland is the default storage location unless a contract says otherwise, with transfers to the United States under adequate controls. Case dispatch to agencies and beneficiary institutions worldwide makes the service cross border by nature.

A partner inquiry form is the only entry, with no price for either product. An intelligence feed and a staffed victim response service differ in kind, so the missing basis of charge matters twice over.

Banks, crypto platforms and cyber insurers each get an industry page, the last sitting where scam losses become claims. Subject coverage stays narrow, authorized payment scams and the mule networks receiving the money, and no institution is named in any segment.