Our Services Contact Us Newsletter

Goodix Combines Fingerprint Matching and Secure Elements for AI Devices

Abstract illustration of a fingerprint with concentric blue rings depicting biometric sensor technology

Goodix has introduced a fingerprint authentication design that pairs matching inside the fingerprint module with an embedded secure element that stores credentials for connected and AI-enabled devices. It is the top tier of a new three-level portfolio that also includes Match-on-Host and Match-on-Chip fingerprint options for makers of AI phones and PCs, wearables, robots, smart-home products and vehicle cockpits.

The design joins Goodix’s Match-on-Chip technology, or MoC, with an embedded secure element, or eSE. Fingerprint image capture, feature extraction, template storage and matching take place within the module. The secure element provides a separate protected location for credentials, including application programming interface keys and OAuth tokens used to authorize access to software services.

Goodix says that arrangement can link a successful fingerprint check to permission to use a credential. In an AI device, for example, the user could be required to authenticate before the equipment performs an action governed by the stored authorization.

Goodix’s Match-on-Chip modules comply with the Secure Device Connection Protocol (SDCP) standard and use unique per-device keys to help protect fingerprint data and communications with the host processor. In the combined design, the module sends match results to the secure element over an encrypted channel, so the secure element can confirm that the legitimate user authorized the interaction. Goodix says the secure element chip is certified to internationally recognized security standards.

The integrated offering supports FIDO2 authentication and TLS 1.3 communications, according to Goodix. FIDO2 uses cryptographic credentials for authentication, while TLS protects data exchanged over a network connection. When a device accesses a cloud-based large language model, Goodix says credentials pass from the secure element over an encrypted connection without being exposed in plaintext to the host processor.

The portfolio covers capacitive, optical and ultrasonic fingerprint sensors. Goodix’s ultrasonic sensor, which first shipped at scale in the Vivo X100 Ultra smartphone in 2024, is being offered for screenless IoT and AI devices, where users press a designated area on the device cover to unlock it, including with wet or oily fingers.

The secure element can also enforce policies such as requiring fingerprint authorization for high-risk operations, and it keeps tamper-resistant audit logs that Goodix says are meant to prevent AI agents from altering the records. Goodix cites an IDC forecast that China’s market for secure AI agent applications will reach $1.6 billion by 2028.

Sources: Goodix

–

By the ID Tech Editorial Team