Quote and proposal → Identity record check → Selfie or video session → Face match → Underwriting → Policy bound
Every step can succeed on manipulated media. The authenticity question is never asked.
Insurance / NeuroVerify
Deepfake Detection API for Insurance Companies
Insurers that sell policies remotely rely on a selfie or video session to confirm the proposer is real and present. NeuroVerify checks that media for deepfakes, face swaps and injected video, and returns one risk score with evidence before cover is bound.
It fits the selfie or video step your onboarding already runs: no new step for the customer, one API call on the media you already capture.
NeuroVerify does not verify identity documents, run your video identification sessions or make underwriting decisions. It adds authenticity evidence; your policy decides.
Powered byNeuroVerifyExplore the product
- #1 of 36on the public VendorBench-100
- <200 msp95 on eligible image requests
- 6 → 1six checks, one unified score
- 0new steps for your customer
The gap
A policy can bind while the proposer was never actually there.
Remote insurance onboarding already has real controls: identity record checks, a selfie or live video session, face matching against the identity on file, and underwriting rules. In many markets, regulators now accept video-based identification in place of a branch visit for higher-value policies.
Each control answers a narrow question. Record checks confirm an identity exists. Face match confirms two images look alike. Underwriting rules price risk. None of them asks whether the face on screen was captured from a real person at that moment, or generated, swapped or replayed.
Why liveness alone misses it
Most liveness checks are built for physical presentation attacks: a printed photo, a mask, a screen at the lens. A deepfake stream injected below the camera layer never presents anything physical, so there may be nothing for that check to object to. Real-time face-swap tools now run on consumer hardware, and a clean-looking video session carries less weight than it used to.
The cost of attempting this fraud keeps falling while remote sales volumes grow. That is the exposure a deepfake detection API for KYC is meant to cover.
Threats
How impersonation reaches a bound policy.
These are attack techniques seen across remote onboarding, not claims about any specific insurer.
01
Live face swap in a video session
The attacker joins a video identification call with a face-swap pipeline. The synthetic face follows their speech and head turns, so prompts are answered and the face on screen belongs to nobody on the application.
02
Generated selfie on the instant path
Policies that bind on an identity check plus a selfie depend on one frame. An AI-generated or swapped selfie arrives in the same format as a genuine one and can pass a face match built from the identity photo.
03
Agent-assisted identification fraud
A third party completes the selfie step for someone who cannot pass identification themselves, such as a person previously declined or with no insurable interest. The documents may be genuine; the person at the camera is not.
04
Pre-recorded proposer video
A prepared clip is played into the session in place of a live feed. It is easier where the workflow accepts uploaded video rather than a fully interactive call.
05
Virtual camera injection
The attacker replaces the camera feed at the driver or device layer. The app receives a well-formed stream that never came from a lens, which physical-spoof checks may never inspect.
06
One synthetic face, many policies
The same manipulated face reappears across applications with different names, nominees or agent codes. Each file looks acceptable alone when reviewed in isolation.
Background: how face swaps are produced, synthetic identity fraud and virtual camera injection.
Claims media is a different workflow from onboarding. See Insurance Claims Media Evidence Review for that case.
Where NeuroVerify fits
One authenticity check inside the onboarding step you already run.
NeuroVerify sits between media capture and your face match and underwriting logic. It works on the proposer selfie or video frames you already collect, and leaves every existing identity control with its owner.
Quote and proposal → Identity record check → Selfie or video capture → NeuroVerify → Face match → Risk score and underwriting policy
One call is added and nothing is removed. Your provider keeps identity comparison and session conduct.
What runs in one request
NeuroVerify checks the face image, video or audio for deepfakes and manipulation beyond liveness. All six signals run together and resolve into one risk score:
- Visual Safety Screening - Live Presence Verification - Face Manipulation Detection - AI Generative Model Analysis - AI Pixel-Level Manipulation - AI Content Compliance
The response carries the unified score plus signal-level evidence. Your own policy makes the final decision.
What stays with you
Your identity vendor, your video identification platform, your underwriting engine, your agent and partner channel controls, your consent and retention rules, and the decision on every proposer. See Insurance Customer Onboarding Review for where the control fits.
NeuroVerify Unified Trust LayerSix parallel analysis layers
- Visual Safety Screening
- Live Presence Verification
- Face Manipulation Detection
- AI Generative Model Analysis
- AI Pixel-Level Manipulation
- AI Content Compliance
Send the proposer selfie or frames
Pass the image or video frames your journey already captures. No change to the customer experience.
Call one API
Use the REST API or an SDK from your orchestration layer, mobile app or backend.
Read the unified risk score
Get one score with signal-level evidence for the policy file.
Route by your policy
Continue to quote, send to manual underwriting or block, under your own thresholds.
What you get
What insurers get from adding the check.
Fewer policies bound to impersonators
Manipulated selfies and swapped video are flagged before cover is issued, not found at claim time.
One evidence record per file
The risk score and signal evidence attach to the policy file for later fraud review or audit.
Review effort goes where risk is
High-risk results go to manual underwriting; clean results continue without added friction.
Coverage across channels
The same call works for direct digital, agent-assisted and video-session journeys.
No new customer step
The check runs on media you already collect, so conversion is not asked to pay for it.
Questions
What teams ask before they integrate.
We already run liveness. Do we still need NeuroVerify?
Liveness answers whether a live person appears to be in front of the camera. It was built for presentation attacks such as a printed photo, a mask or a screen held to the lens. It was not built to tell whether the stream reaching your server came from a camera at all, or whether the face inside it was generated or swapped.
NeuroVerify runs inside the selfie or video step you already have. Live Presence Verification is one of its six signals, and the other five look for face manipulation, generative-model traces, pixel-level edits and unsafe content in the same request. That removes the seam between two products where impersonation attacks tend to land. You can keep your current liveness vendor during an evaluation and compare outcomes on your own media. See deepfake detection vs liveness detection.
Does it replace our identity vendor or video identification platform?
No. NeuroVerify adds authenticity evidence next to those platforms. Identity record checks, session conduct and regulator-facing process design stay where they are.
Where in the journey does the call belong?
After the selfie or video frames are captured and before face match and underwriting rules act. That way authenticity evidence is available when your existing provider compares faces.
How hard is it to integrate?
NeuroVerify offers native mobile SDKs (Android/Kotlin, iOS/Swift, Flutter), backend SDKs (Python, TypeScript/Node.js, Go, Java) and a REST API. Integration is two lines of code, and teams can be live in under 24 hours.
The same API accepts selfie images from instant journeys and frames from live video sessions, so one integration covers both channels. See integrating into an existing KYC stack.
How is our data handled, and where can it run?
NeuroVerify analyses the face images, video or audio you send and returns a risk score with evidence. NeuroVerify does not validate identity documents or approve applications. Consent, retention and residency stay under your policy. See product data handling for how media is treated.
Deployment options include the managed cloud API, private cloud or customer VPC, a secure container, on-premises and air-gapped environments. Neural Defend holds ISO/IEC 27001, ISO/IEC 27701, ISO/IEC 42001 and SOC 2 Type II certifications, shown on the trust page.
How fast is it, and will it slow the journey?
NeuroVerify targets under 200 ms at p95 on eligible image requests under defined test conditions. The test methodology is available on request, so you can check it against your own media sizes and network path.
It is one API call on media you already capture, so it adds a single request rather than a new user step. Latency for video depends on clip length and your deployment, so measure it in your evaluation.
How do we handle false positives and set thresholds?
NeuroVerify returns one unified risk score and signal-level evidence. It does not approve or reject anyone. Your policy sets the thresholds and decides what each band does.
Many teams route high-risk results to manual underwriting rather than automatic rejection, so an underwriter sees the evidence next to the rest of the application. See Synthetic Identity Application Investigation.
Can we test it on our own media first?
Yes, and that is the test that matters. Public results help: NeuroVerify ranked first of 36 systems on VendorBench-100, with a detection score of MCC 0.876 against 0.646 for the next-best vendor. But your capture devices, compression and attack mix decide real performance.
Use authorized samples from direct, agent-assisted and video-session paths, plus your current decision states. Compare scores and evidence against your current decisions. For a vendor shortlist, see the top 5 deepfake detection APIs.
Evaluate NeuroVerify
Test NeuroVerify on your insurance onboarding media.
Bring authorized media from your own journey. NeuroVerify returns a risk score with signal-level evidence, so you can compare it with your current checks before changing anything.