Brave Leo adds E2E encryption, Oblivious HTTP and verifiable confidential computing for AI chats
gnukeith · x · 2026-10-16
Brave announced that Leo, its browser AI assistant, now offers verifiable privacy: conversations with the open-weight GLM model are protected by end-to-end encryption and hardware-backed confidential computing.
- The browser encrypts each prompt on-device so only an attested Trusted Execution Environment (TEE) can decrypt and run inference
- The TEE must present hardware-signed attestation verifying the secure hardware, model integrity, and publicly auditable code
- The stack includes Oblivious HTTP and Verifiable Confidential Computing, framed as "privacy you can verify, not just trust"
Related event: Brave Leo Adds GLM with End-to-End Encrypted, Verifiable AI Chats(2 posts)→
More from Safety
- Critique: interpretability research keeps running intuitions ahead of rigorous evaluation — dhadfieldmenell · 2026-10-16
- Frontier Red Team member: AI worms are dumb today, but the exponential is unsettling — logangraham · 2026-10-16
- Paper warns of "agentic flooding": AI tools are straining courts worldwide — TuhinChakr · 2026-10-16
- NYC comptroller overseeing $325B pension asks Anthropic to delay IPO over safety concerns — rohanpaul_ai · 2026-10-16
- Wired: AI Hacks Leave a Mess That AI Companies Aren't Paying For — Wired AI · 2026-10-16
- Researcher: alignment's basic problems are known, yet almost nobody works on them directly — jd_pressman · 2026-10-16