doxx.net has raised $38 million to support its ADN platform, which is designed to prevent AI agents from causing harm while acting on a user's behalf, according to SecurityWeek.
The platform targets what the company calls 'agentic misadventure' — unintended or harmful actions taken by an AI agent while it operates under the user's authority. ADN is meant to stop these problems while the agent is actively working.
The funding round reflects growing investor interest in safety controls for autonomous AI systems.