According to monitoring by Beating, Liquid AI has released ShieldFlow, a local privacy data de-identification tool. ShieldFlow supports the local operation of the built-in Liquid Foundation Model, which can automatically identify and mask sensitive data on the device, preventing privacy information from being sent to the cloud. During operation, ShieldFlow does not require GPU computing power and has very low memory usage, allowing it to run offline in real-time on the vast majority of standard personal computers. ShieldFlow was showcased as part of the Foundry Local program at the recent Microsoft Build 2026 Developer Conference. Additionally, at the Taipei International Computer Show (COMPUTEX 2026), ShieldFlow was demonstrated running on AMD laptops. Liquid AI has now opened early access applications for ShieldFlow.
All Comments