
900 attacks. Zero got through.
That is the result from Incode’s iBeta Level 3 evaluation: a comprehensive assessment of our liveness detection algorithm across both iOS and Android platforms. The test, conducted by an independent and accredited testing laboratory, simulates the most sophisticated presentation attacks available today, and Incode stopped every single one.
This is a big deal. iBeta Level 3 is the highest standard in the iBeta Presentation Attack Detection (PAD) testing framework. Achieving 100% across both major mobile operating systems is a rare result.
The iBeta PAD testing framework evaluates biometric systems against presentation attacks, attempts to defeat liveness detection using 2D images, 3D masks, video replays, deepfakes, and other spoofing materials. Level 3 uses the most adversarial and realistic attack scenarios in the iBeta framework.
Passing at Level 3 across both iOS and Android means our algorithm demonstrated consistent, platform-independent performance under controlled, high-rigor conditions.
This is not a marketing certification. It is a technical pass/fail test with publicly verifiable methodology. The results are independently verifiable through iBeta.

.png)

The fraud environment that our customers navigate is not stable. Presentation attacks are becoming more accessible and more realistic. Freely available tools now make it possible for low-sophistication actors to attempt spoofing at scale. The threat model has shifted, and our evaluation standard has to reflect that.
We submitted to iBeta Level 3 because it is the most demanding independent evaluation available. If our system can pass at Level 3 under controlled adversarial conditions, it is better positioned to hold up under real-world attack conditions.
This also matters for our customers in regulated industries. Some verticals are moving toward requiring Level 3 conformance, or are already treating it as a qualifying criterion for vendor selection. Achieving it now means our customers are ahead of that curve.
iBeta Level 3 is a milestone, not a ceiling. The threat landscape will continue to evolve, and our evaluation and testing cadence will evolve with it. We continue to run internal red-team adversarial testing, monitor emerging attack tooling, and update our models accordingly.
We’re also continuing to push on accuracy and throughput. Liveness detection that stops attacks but creates friction is not a complete solution. The goal is high accuracy, low false rejection, and seamless user experience, simultaneously.
Incode was named a Leader in the 2025 Gartner® Magic Quadrant™ for Identity Verification. Download the report.