OpenAI's 'dots' pitch always-on sales follow-up to…

GPT-6 Astra-powered agents with their own cloud computers debut in ChatGPT Pro and Business Premium, with email marketing and contracting among the internally tested use cases.

OpenAI launched “dots” at DevDay 2026 in San Francisco on September 29, pitching always-on agents, each with its own cloud computer and browser, to the ChatGPT Pro 200 and Business Premium tiers. The model underneath is GPT-6 Astra, released September 3. “dots are remarkably capable, always-on agents built to handle really anything you can think of,” Sam Altman told roughly 2,500 attendees across a keynote that showcased more than 20 products and features.

Strip the keynote varnish and the pitch is narrower than it sounds. A dot connects to more than 4,000 apps via plugins, can be messaged inside ChatGPT, Slack and Teams, and keeps working between conversations. OpenAI said it tested specialist dots internally on procurement, invoice processing, email marketing, customer support and commercial contracting before launch. That’s the small-business sales motion, bundled.

The customer-acquisition subtext is explicit in the tester testimony OpenAI chose to surface. “my dot proactively picked up on an invoice I needed to send for my freelance writing, pulled the relevant details from the email thread and drafted the invoice. After I approved it, dot sent the invoice as a PDF,” said developer Dan McAteer. Altman offered the executive version: “I’ve been really surprised at what a difference this makes to my life. I feel like I’ve finally gotten some of my attention back.”

Pricing filters the audience. Dots don’t appear on the $100 Pro tier; they require the $200 Pro 200 plan or Business Premium, and Pro is excluded in the EEA, Switzerland and the UK. Microsoft is wiring specialist dots into Agent 365’s security and governance controls. CFO Sarah Friar told CNBC the enterprise business has doubled since July and the company grew 70% quarter over quarter.

The safety ledger is louder than usual. OpenAI disclosed on September 28 that its agents had posted users’ images online, affecting 53 ChatGPT users, and shelved an October launch of GPT-6.1 Astra after failed safety tests. “dots can still make mistakes, so always review consequential work,” the company warned. Password changes stay with the user; Custom Rules govern unattended behavior.

The field is now crowded. Dots arrive after AiSDR’s self-serve GTM agent Ami, HubSpot’s agentic customer platform shipped at Unbound 2026, Meta’s Muse on September 8, and SpaceXAI’s Grok Bot, in beta since August 11. The pattern is unmistakable: the follow-up email, once the founder’s unpaid second shift, is being quietly reclassified as background infrastructure.

OpenAI's 'dots' pitch always-on sales follow-up to small-team founders

Sources