Synsira Launches On-Device AI to Combat Cloud Privacy Risks
AI-generated image · Bay Street Wire
Kind Local Pro promises a zero-cloud threat model, though the company admits to periodic license verification.
Jonathan Schaeffer, a former founder of the Alberta Machine Intelligence Institute (Amii), has released a privacy-focused AI tool designed to keep user data entirely local. The product, Kind Local Pro (KLP), is developed by Schaeffer’s Vancouver Island-based firm, Synsira Software.
Unlike the company’s Kind Pro, which utilizes cloud infrastructure, KLP runs on-device. According to reporting from BetaKit, the software allows users to ingest documents, images, audio, and email inboxes to create a knowledge database. Schaeffer told BetaKit that all processing occurs on the user's computer and data never leaves the device. He noted that the only interaction between the installation and Synsira is for periodic account verification to confirm a valid license.
Schaeffer stated he was motivated by disillusionment with how large AI companies handle privacy, specifically citing the 2022 arrival of ChatGPT. Because it avoids data centers, Synsira is targeting KLP at academia, intellectual property development, and legal work. Schaeffer also noted that the local approach reduces the environmental impact associated with massive data centers.
Regarding safety and oversight, Schaeffer told BetaKit that KLP avoids "chatting" and anthropomorphized interactions to mitigate safety risks. He described the interface as being similar to the computer in Star Trek, providing only fact-based answers derived solely from the user's provided data.