AI · Official
OpenAI Shares How It Secures Always-On OpenClaw Agents
OpenAI Developer changelog (mirror) ·

At a glance
OpenAI explained how it keeps OpenClaw, an AI agent that runs 24/7, safe for use inside the company. Hundreds of OpenClaw agents are already running across the firm.
Key points
- Agents can change their own settings and write code to reach a goal, so they can be risky.
- OpenAI treats the settings side as trusted and the AI-running side as untrusted, and keeps them apart.
- Each agent's access to internet addresses, files, programs, and passwords is limited.
- Agents run in separate containers, and the fully unrestricted access mode is blocked.
Why it matters
It offers a real example of how companies can put safety limits on AI agents that act on their own.
Original source
Securing persistent agents like OpenClaw at OpenAI
OpenAI Developer changelog (mirror) · Nov 12, 09:00 KST
Read on OpenAI Developer changelog (mirror) ↗AI summarized the original article. Check the original for accuracy. Copyright belongs to OpenAI Developer changelog (mirror).
More from this field
- Google Brings Duplex Calling to Gemini, Letting AI Phone Businesses for You
Google has moved its Duplex phone-calling technology into the Gemini app, so Gemini can now call businesses for you. The system has changed a lot since 2018, shifting from narrow custom models to a single stack built on Gemini.
Google Research blog · Nov 21, 08:30 KST - Google Releases Project Genie, an AI That Generates Explorable Worlds
Google has released Project Genie, an AI tool that creates interactive worlds you can explore. The video shows how to access it, walks through a generated world and a T-Rex demo, and discusses why this early technology matters.
The AI Advantage · Nov 21, 08:21 KST - Anthropic Employee Says Claude Helped Prove Riemann Hypothesis
An Anthropic employee and mathematician, Levent Alpöge, says he used the company's Claude chatbot to create a detailed proof of the Riemann hypothesis over about six weeks. The proof has not yet been checked, and it is over 250,000 lines long in Lean, a computer language used to confirm maths proofs.
BBC News · Nov 21, 08:08 KST - Cole Medin Live-Streams His AI Software Factory Turning Ideas Into Working Software
Cole Medin runs his AI software factory live, taking rough requests from viewers through planning, building, and checking until working software runs on screen. He explains how it works and what he learned building it. The stream also shows a game the system built.
Cole Medin · Nov 21, 08:08 KST