Research

How we think about building a personal agent that acts on your behalf — and keeps that autonomy verifiable, safe, and legible. Notes, theses, and evaluation work as we go.

Identity & Security·July 15, 2026

The Identity Layer of AI Agents

A research overview of how AI agents prove who they are and what they may do — the standards, the vendors, a real breach, and the gap between authorization and verified intent.

Read

Autonomy & Safety·June 6, 2026

Mae Control Research

A thesis and evaluation framework for keeping an autonomous personal agent verifiably aligned with what a person actually intended.

Read