Product definition and guardrails

VoxEQ Fraud Screen is a Risk-Based Assessment (RBA) capability for inbound contact centers. It provides an early-call risk signal intended to help regulated organizations decide how much scrutiny is appropriate for a given caller—especially when the caller is first-time or infrequent, and when enrollment-based methods are impractical.

What Fraud Screen is

• An upstream risk signal derived from real-time analysis of voice bio-signals.

• Built for non-face-to-face servicing environments where proportional controls are expected.

• Designed to help teams apply risk-based controls consistently before sensitive actions are performed.

What Fraud Screen is not (critical positioning)

Fraud Screen must be treated as decisioning input, not as a control that “proves identity.”

• Not Identification/Verification (ID/V).

• Not caller authentication.

• Not a replacement for knowledge-based checks, out-of-band verification, or other authentication steps.

In other words: Fraud Screen answers “How risky does this call look?”—not “Is this person truly the account holder?”

Where Fraud Screen sits in the call flow (and what it informs)

Fraud Screen is designed to operate upstream of ID/V, early enough to influence routing, limits, and step-up decisions.

Typical inbound flow

1. Incoming call connects (IVR / virtual agent / live-agent queue).

2. Fraud Screen runs early and returns a risk signal.

3. Contact center applies policy-appropriate handling:

4. Low-risk → proceed with standard handling.

5. Elevated-risk → apply proportional friction (step-up) and/or tighter controls.

Actions Fraud Screen can inform

Fraud Screen is most valuable when paired with predefined policies that map risk levels to concrete controls.

Important: Fraud Screen informs decisions like step-up, limits, and escalation; it does not itself “approve” or “deny” identity.

Ideal use cases (high-risk + low-frequency) where RBA matters most

Fraud Screen is designed for scenarios where the “caller is unknown enough” that enrollment-based security is brittle, but the interaction is still sensitive.

Examples:

• Banking / credit union servicing for infrequent callers (e.g., retirement accounts, beneficiary changes)

• Insurance claims and policy changes

• Healthcare payer/member eligibility or billing inquiries

• Government benefits servicing

• Utilities start/stop service and high-risk account changes

• Complaint desks / escalation hotlines where social engineering attempts concentrate

Privacy and data-protection posture (what Fraud Screen explicitly avoids)

Fraud Screen is designed to reduce privacy overhead and compliance risk by avoiding biometric-retention workflows.

• No voiceprints

• No biometric enrollment

• No recordings

• No stored files

• No back-office data handling

• No workflow changes required (policy enforcement can be implemented in the existing contact-center logic)

VoxEQ also publishes an AI ethics statement describing privacy and responsible-use commitments (for example: not attaching personal identifiers to biometric data; providing labels/risk scores rather than exposing biometric information). See: VoxEQ AI Ethics Statement.

Why now (U.S. 2024–H1 2025): fraud trends that increase the value of upstream RBA

The case for an upstream, coverage-first risk signal in the voice channel is stronger when:

1) scams are driving high losses, 2) mule networks and real-time rails accelerate cash-out, 3) account takeover pressure grows, and 4) fraud tools shift toward remote access and stolen-device execution.

Breach volume as a caller-impersonation enabler (context) The Identity Theft Resource Center (ITRC) counted 3,158 data compromises in 2024 (vs. 3,205 in 2023 and 1,801 in 2022). As breach exposure accumulates, more personal data becomes available to support impersonation and social-engineering in inbound servicing flows. Examples frequently cited in breach reporting include large-scale exposures such as NYU applicant data (3M+) and Community Health Center data (1M+).

Terms clarification (mini‑glossary): fraud vs. scams vs. breaches

To keep “fraud trends” discussion precise, it helps to separate three related (but different) concepts:

1. Traditional identity fraud: misuse of someone’s identity or account (e.g., account takeover, new-account fraud) where the criminal is acting as the victim to obtain value.

2. Scams / social engineering: deception that manipulates a person (or an agent) into authorizing an action (e.g., changing details, moving funds, resetting credentials) even when some “identity checks” appear to pass.

3. Identity theft via unauthorized data access / breaches: the upstream enabler—unauthorized access to data that supplies attackers with enough accurate details to sound legitimate during inbound calls.

0) Identity fraud + scams reached $47B in 2024 (AARP × Javelin)

A 2024 report produced by Javelin Strategy & Research and co-sponsored by AARP estimates $47B in total losses to identity fraud and scams in 2024 (up $4B vs. 2023), based on a survey of 5,023 U.S. adults. It also splits losses into:

• $27B from traditional identity fraud, affecting 18M people

• $20B from scams

• Account takeover (ATO): $15.6B in losses

• New-account fraud: $6.2B in losses

The same summary notes that scammers’ first contact is increasingly via text message (54%, up from 49% in 2023), and that 71% of those who reported scam losses also reported sharing personal information.

Source: AARP summary of Javelin’s 2024 identity fraud study.

Why this matters for Fraud Screen: when loss totals are this high—and when scam-driven fraud pushes victims (or impostors with victim data) into inbound servicing flows—contact centers need an upstream way to decide how much scrutiny is appropriate before completing high-impact requests. Fraud Screen is designed to provide that early-call, coverage-first RBA signal, especially for first-time and infrequent callers where enrollment-based controls don’t apply.

1) Investment scams are a major loss driver (calendar year 2024)

The FBI’s 2024 Internet Crime Report (IC3) reports that victims of investment fraud reported losses of over $6.5B in 2024, and that overall reported losses exceeded $16B (released April 23, 2025). Source: FBI press release on the 2024 IC3 report and the underlying PDF: 2024 IC3 Internet Crime Report (PDF).

Why this matters for Fraud Screen: scam-driven fraud often culminates in inbound servicing attempts (policy changes, credential resets, payment instructions). An early RBA signal helps ensure proportional friction is applied before the request reaches a permissive workflow.

2) Mule-account and money laundering indicators increased sharply (H1 2025 vs H1 2024)

BioCatch reported that U.S. financial institutions saw a 168% spike in detected money laundering accounts, and that confirmed money-laundering cases more than doubled in H1 2025 vs H1 2024. Source: BioCatch press release (Sept. 9, 2025) and BioCatch’s report page: 2025 Digital Banking Fraud Trends in the United States.

Why this matters for Fraud Screen: organized scam operations require mule accounts and rapid movement; contact centers handling “one-time” calls (life events, emergencies) are a common weak point. Upstream RBA improves coverage for callers who are not enrolled and not “known enough” for other controls.

3) Impersonation and purchase scams remain common, while real-time rails speed cash-out

BioCatch notes impersonation and purchase scams as dominant by volume, and highlights stablecoins plus authorized push payments (APP) as key real-time movement mechanisms. Source: BioCatch press release (Sept. 9, 2025).

Government + business impersonation scams: losses, new rules, and early enforcement (FTC, 2024)

Impersonation is not just a “fraud trend” — it is a large loss category with active U.S. regulatory attention. PYMNTS reported (citing the FTC) that government and business impersonation scams were consistently among the most-reported fraud types and drove $2.95B in consumer losses in 2024. PYMNTS also notes the FTC’s Government and Business Impersonation Rule took effect in April 2024, and that post-rule enforcement included five cases plus the takedown of 13 websites accused of illegally impersonating the FTC online — including action involving Superior Servicing, which allegedly posed as affiliated with the U.S. Department of Education to extract money from student-loan borrowers. Source: PYMNTS coverage of FTC impersonation scam figures and enforcement.

Why this matters for Fraud Screen: impersonation attempts frequently arrive as inbound calls requesting high-impact actions (account changes, payout instructions, credential resets). A coverage-first RBA layer helps teams identify elevated-risk calls early — including first-time and infrequent callers — and apply proportional step-up before the call enters a permissive workflow.

Scam victimization spans demographics (not just older adults)

The same PYMNTS article cites PYMNTS Intelligence research that 30% of Americans (about 77 million people) reported scam losses in the last five years and that victims span all demographics. Source: PYMNTS coverage.

Why this matters for Fraud Screen: if victimization isn’t confined to a single “high-risk” demographic, controls that depend on prior enrollment or “known” customers will leave gaps. Fraud Screen is designed to provide upstream, enrollment-free RBA coverage for every caller, so policy-based step-up can be applied consistently when the call looks riskier. Why this matters for Fraud Screen: when cash-out becomes faster, “detect later” becomes “detect too late.” Fraud Screen supports faster decisioning about when to:

• add step-up friction,

• slow down risky servicing,

• route to specialized teams, or

• apply temporary limits until verification is completed.

4) ATO pressure is rising while fraud execution techniques become more operationally scalable

BioCatch reported (H1 2025):

• Account opening (AO) fraud attempts down 18%

• Attempted account takeover (ATO) up 13%

• Attempted ATO using remote access trojans (RATs) up 50%

• Stolen-device fraud increasing

• For credit unions, nearly 20% of fraud relates to card activity

Source: BioCatch press release (Sept. 9, 2025).

Why this matters for Fraud Screen: ATO and RAT-enabled fraud often aims to defeat “traditional” identity checks with speed, persuasion, and operational playbooks. Fraud Screen adds an upstream RBA layer that helps teams apply defense-in-depth in the voice channel without forcing enrollment.

How to deploy Fraud Screen conceptually (policy-first)

A practical way to operationalize Fraud Screen is to treat it as an input to a contact-center risk policy:

• Define risk tiers (e.g., low / elevated / high) and what each tier means.

• Pre-approve step-up and limitation actions for each tier.

• Make outcomes deterministic (so agents and auditors can understand decisions).

This preserves the intended role of Fraud Screen: early, privacy-preserving RBA coverage, upstream of ID/V.