OpenAI launches dots agents on GPT-6 Astra with user sign-off for sensitive tasks
In brief
- OpenAI introduced dots, autonomous agents on its GPT-6 Astra model, on September 29, 2026.
- Dots agents are built to work continuously across Slack and Microsoft Teams.
- Dots needs user sign-off before sensitive actions like password changes or data deletion.
- Meta's Muse agent was released earlier in September, Crypto Briefing reported.
- Codex and ChatGPT Work have more than 35 million weekly users, OpenAI reported.
An agent that asks first
OpenAI's new agents don't run entirely unchecked. According to Crypto Briefing's report, dots has to get the user to sign off before it carries out a sensitive action, with password changes and data deletion given as the examples.
That's the consent step.
It sits alongside a wider set of controls that both companies are leaning on. OpenAI and Meta rely on custom data controls, adjustable user permissions and settings that keep enterprise usage data from being retained, the outlet reported. For workplace apps that live inside Slack or Teams all day, those settings carry real weight (an agent that keeps working in the background is also one that keeps touching company data).
Meta's Muse and OpenAI's user base
Meta isn't sitting this one out. The company released its Muse agent earlier in September 2026, and Crypto Briefing said it quickly racked up millions of downloads.
OpenAI's own figures are on a different scale. As of late September, the company reported more than 35 million weekly users across its Codex and ChatGPT Work products, according to Crypto Briefing. OpenAI also said its consumer ChatGPT app draws over 1.2 billion weekly users.
Trust, and the disclosure debate
Crypto Briefing framed the focus on user trust as following earlier incidents in which AI models interacted in ways people didn't expect. It's the outlet's read on why both companies are putting permissions and data controls up front.
The report also placed the rollout within a continuing debate over safety disclosures, one that centers on how AI systems behave and how much decision-making they handle on their own. Dots lands right in the middle of that question: it's built to keep running across workplace tools, yet it's designed to stop at the riskiest actions.
So what happens when an always-on agent reaches a task like deleting data? According to OpenAI, it waits for the user to sign off.
Frequently asked questions
What are OpenAI's dots agents?
Dots is a set of autonomous agents that OpenAI introduced on September 29, 2026, running on its GPT-6 Astra model, according to Crypto Briefing. They're designed to keep working continuously across workplace platforms, including Slack and Microsoft Teams.
How does dots handle sensitive actions like deleting data?
Before dots carries out a sensitive action, such as changing a password or deleting data, it needs the user to sign off, Crypto Briefing reported. OpenAI and Meta also rely on custom data controls, adjustable user permissions and settings that keep enterprise usage data from being retained.
