PLATFORM • VERIFYHUMAN

AI-resistant respondent verification.

Multi-signal liveness, pre-survey fraud prevention. Built for the post-bot reality where 99.8% of AI bots pass traditional attention checks. Embedded in RIWI's Research and Audience by default. Standalone for any external survey or system.

Every survey is now a Turing test.

01 / 03

31%

Fraudulent Responses

NORC + CloudResearch, 2026

02 / 03

99.8%

AI Bot Evasion

Of recent AI bots pass traditional attention checks (Dartmouth Westwood, 2025)

03 / 03

38%

Discard Rate

When verification is applied post-hoc (Kantar)

THE PROBLEM

AI bots have already broken survey-based research.

In independent testing, 99.8% of recent AI bots passed traditional attention checks — the kind every survey platform has used for a decade. CAPTCHAs are decorative. Behavioral checks are guessable. Post-hoc data cleaning catches some of it, at the cost of 38% discard rates and weeks of doubt about which findings are real.

The verification industry hasn't caught up. The AI bots are improving faster than the defenses.

Read the data quality story

THE SOLUTION

AI-resistant. Pre-survey. Multi-signal.

VerifyHuman is a verification product built for the post-bot reality. It runs before a respondent answers a survey question — pre-survey fraud prevention, not post-hoc cleaning.

It uses multi-signal liveness — combining behavioral, device, session, and biometric signals — to identify the bots that look human enough to fool everything else. AI-resistant by design: the signals VerifyHuman checks are the ones AI bots can't yet fake reliably, and the system updates as the threat surface changes.

Pre-survey, not post-hoc
Multi-signal liveness — behavioral + device + session + biometric
AI-resistant by design
Embedded in RIWI products by default
Standalone for any external system

WHAT GETS CHECKED

Four signals. One verdict.

01 / 04

Behavioral

Movement, hesitation, scroll patterns, response timing. The signals AI agents struggle to fake naturally.

02 / 04

Device

Device fingerprinting, hardware signals, network characteristics. The data trail every device leaves.

03 / 04

Session

Session continuity, prior verification history, geographic consistency. Catches the patterns that cross-survey fraud creates.

04 / 04

Biometric

Liveness checks. Selective, privacy-aware, applied where the signal is needed and the consent supports it.

TWO WAYS TO USE IT

Embedded in RIWI, or standalone anywhere.

01 / 02

EMBEDDED — On every RIWI study by default

When you use RIWI's Research or Audience, VerifyHuman runs underneath. Pre-survey verification on every respondent. Nothing to integrate, nothing to configure.

Explore the Audience→

02 / 02

STANDALONE — On any external survey or system

SDK for major survey platforms. REST API for custom systems. Web interface to configure verification flows. Use VerifyHuman on Qualtrics studies, your own platform, or any third-party system.

Get integration access→

WHY THE OLD APPROACH FAILS

CAPTCHAs were designed for an older internet.

CAPTCHAs were built when bots were obvious. AI defeats them. ReCAPTCHA v3 issues quality scores; AI agents can produce sessions that score well. Attention checks ("select 'somewhat agree' for this question") were designed to catch inattention, not adversarial behavior. AI bots can read instructions.

The category is moving on. Multi-signal liveness combines what each signal alone can't catch — adversarial bots that pass any single check, fail when checked across four dimensions simultaneously. AI-resistant means designed-against the actual current adversary.

This isn't a CAPTCHA. It's the verification layer that comes after the CAPTCHA stops working.

COMPARISON

Traditional defenses weren't built for AI-era fraud

CAPTCHAs and device fingerprinting catch the easy cases. Multi-signal liveness catches what they miss.

Capability	reCAPTCHA	Device Fingerprinting
AI / bot respondents
Survey farms		partial
Duplicate participation
Demographic fraud
LLM synthetic respondents
Pre-survey, not post-hoc	partial	partial

HOW TO INTEGRATE

Three ways in, depending on your stack.

01 / 03

SDK — Survey platform SDK

Drop-in integrations for Qualtrics, SurveyMonkey, Decipher, and major survey platforms.

02 / 03

API — REST API

Server-side verification for custom applications. Verify before survey entry; receive structured signals back.

03 / 03

STANDALONE WEB — Web interface

Configure verification flows without code. Generate verification links, embed them in any survey distribution.

ACCESS

Free on the RIWI platform. Custom for standalone.

Free with RIWI

with platform sign-up

Included on every RIWI study

Pre-survey verification on every respondent
Multi-signal liveness
AI-resistant by design
Nothing to configure

Standalone Starter

$0.10

per verification

Up to 10K/month

SDK for major survey platforms
REST API access
Webhook notifications
Email support
Cross-study deduplication

Talk to sales

Standalone Professional

$0.05

per verification

10K – 500K/month

Everything in Starter
Priority support
Custom thresholds
Analytics dashboard
Volume pricing

Talk to sales

Enterprise

Custom

volume pricing

500K+/month

Everything in Professional
Dedicated account manager
Custom SLA
On-premise option
SOC 2 report
White-label SDK

Contact sales

READY TO VERIFY

Stop bots before they enter your study.

Sign up free and use VerifyHuman on RIWI studies today. Or talk to sales about embedding it in your existing survey infrastructure.

Insight Engine

Audience access

Verification