On-device AI threat defense for organizations and their supply chains

Reken's platform runs AI threat detection on users' devices through the Reken Private Core to keep confidential from having to be processed through third party clouds or frontier LLMs.

Reken protects bidirectional communications—both to and from devices and accounts.

The Reken platform is comprised of three main components:

Model Zero, a suite of proprietary models, analyzes incoming data for threats

Trust Sensor verifies human activity and the validity of outgoing communications and transactions.

Reken API
Enables any application to connect to the Reken Private Core.

Local models run in near real time on ordinary devices without GPUs or AI token costs, and can surface zero day attacks and bots without using private data.

Reken, based in San Francisco was founded in 2024 and is venture backed.