Prove plans launch of privacy-preserving biometric binding with new IP

Behavioral signals and contextual information can go a long way in assessing risk and approving authentications. But Prove’s new General Manager for New Market Initiatives Frances Zelazny argues the addition of privacy-preserving biometrics at the identity-binding stage does more than enable the step-up authentication that comes with additional risk. It also sets businesses up to handle the next phase of internet traffic in which AI agents mix with real people, and all of them need to be able to complete transactions.

Prove has created a unique moat with its device intelligence layer, according to Zelazny, leveraging its transaction volume. Those transactions have reached 30 billion a year, giving Prove visibility into the activities of more than 90 percent of active global adult phone users.

This volume allows Prove to position itself as a tool for not only catching fraud, but also identifying legitimate users.

The company’s mission, therefore, is based around enabling valid authentications and transactions, across banks but also gig economy, marketplaces, fintech, crypto, gaming, hospitality and other sectors.

Prove’s device intelligence can provide a higher level of assurance due to the volume of data it holds and the level of intelligence it can provide, Zelazny says, “but still, the phone is not you.”

For those use cases where the highest levels of assurance are needed, combining a biometric factor with the phone intelligence creates a gold standard.

That biometrics-enabled gold standard extends Prove’s utility to use cases like a person getting a new phone, high-volume money transfers and transactions initiated in a place the person hasn’t been before.

Part of Prove’s pitch, however, has long been the privacy-preserving advantage of providing risk intelligence without handling personally identifiable information.

Customers are naturally more hesitant to handle PII with every data breach that is reported and new data protection regulation that passes, let alone biometrics.

“So doing this in a responsible way was really important in order to achieve the ultimate goal and vision that Roger (Desai, Prove CEO) had. Enter the privacy-preserving biometric component to enable that.”

Iterating from experience

The technology that Zelazny is joining Prove to implement is new, she says, drawing on lessons learned in the past and introducing more cryptographic features to create new IP.

Those lessons come largely from Zelazny’s time with BioCatch and as founder of Anonybit.

The new Prove GM suggests the new technology will scale more effectively and be more hardened against attacks.

“At Anonybit I think we did a great job pioneering the space and creating understanding of why privacy-preserving biometrics are important.” The company was starting to scale, but “we recognized certain things that had to be done in order to really meet the demand.”

“Now,” Zelazny says, “we’re able to take those lessons and implement them from day one.”

Prove is on track to meet its previously-announced goal of bringing this new product to market towards the latter part of 2026. The current phase of work involves several design partners, and there is pent up demand for the privacy-preserving biometrics product in development within Prove’s customer base, Zelazny says.

The company already has a full suite of capabilities, including for identity orchestration, that have long been part of the vision the experienced digital ID executive shares with Prove CEO Desai, and she reports initial responses from that customer base suggest they are ready to embrace it.

Other people have joined the Prove team from Anonybit, but Zelazny says the company has built new technology and IP for various cryptographic techniques, including multi-party computation, completely from scratch. Prove has filed for patent protection on this new IP.

That technology goes towards supporting biometric authentication, one-to-one and one-to-many matching for deduplication, velocity checks and blacklist checks, all the way through to downstream authentication and downstream account recovery.

“At Anonybit we just had what we had,” Zelazny says. “Here, we’re able to orchestrate journeys and incorporate silent authenticators.”

She describes Prove’s current technology as a complete three-legged stool, resting on the new biometrics capabilities, along with behavioral signals and other contextual information.

Account recovery to agentic AI

Account recovery is the Achilles heel common to identity lifecycle, or “circle of identity,” Zelazny points out, arguing that this will continue to be the most vulnerable point in any system without “foundational binding from the beginning of onboarding.”

Combining biometrics with Prove Prefill during enrollment enables a range of initial fraud checks, but also binds the biometric to the account and data associated with the phone number in a Prove ID.

During subsequent transactions, silent authentication will protect activities with lower risk, and a selfie biometrics check can meet any step-up authentication requirements in higher-risk situations.

Zelazny also notes potential applications such as for shared devices or non-personal devices like kiosks, replacing OTPs and other legacy authentication approaches.

Binding an individual’s biometric to an AI agent that represents them also allows relying parties to confirm authorization through biometric authentication, guarding against rogue actions or other risk-elevating scenarios.

Businesses are largely unready to handle agentic AI risk, as shown by the hack against Hugging Face by rogue OpenAI agents. The combination of foundational biometrics binding with Prove’s existing technology can help organizations prepare for the next phase of online traffic, Zelazny says.

That next phase represents “a moment of truth” in which those that don’t properly secure agentic traffic “are going to be in big trouble.”

With the addition of privacy-preserving biometrics, Zelazny says Prove is ready to meet that moment.

Article Topics

biometric binding  |  biometrics  |  continuous authentication  |  digital identity  |  Frances Zelazny  |  identity assurance  |  identity orchestration  |  multi-party computation  |  Prove