Imagine an AI that doesn’t wait for you to ask it something. It knows you have a vendor contract renewal coming up, pulls the relevant emails, drafts a counter-proposal, and puts it in your inbox before your morning coffee. That’s what OpenAI is going for with Dots.
Launched this week at DevDay 2026, Dots are autonomous AI agents powered by GPT-6 Astra, OpenAI’s most capable model. Unlike chatbots that respond when prompted, Dots run continuously in the background, connected to over 4,000 apps and services including email, calendars, Slack, Teams, and dozens of business tools. You set a goal, and Dots works toward it while you’re doing other things.
The word “agent” gets overused. Here it actually means something specific. A Dot can browse the web on a dedicated cloud computer, take actions inside connected apps, and adapt its behavior based on what it learns about your preferences over time. OpenAI built in a read-only mode that activates when you’re offline, so the agent gathers information without taking irreversible actions without your approval. That’s a meaningful safeguard.
Right now, access is rolling out to Pro and Business Premium subscribers, with Enterprise access available in beta for organizations whose admins enable it. The pricing reflects the ambition. This isn’t a chatbot subscription.
The competitive pressure behind Dots is real. Meta’s Muse has been gaining traction in the consumer space, running across Facebook, Instagram, and WhatsApp with massive built-in distribution. OpenAI is countering with deeper integration and a more professional use-case focus. The race to own your daily workflow just got much more serious.
What does this mean for a small or mid-sized business? Three things worth thinking through.
First, the productivity math changes. If a Dot can handle calendar coordination, draft outbound communications, and monitor specific workflows without requiring constant human direction, the question becomes how you structure your work to take advantage of that. Not every task needs to be rethought. Many do.
Second, the integration question gets serious. Dots connects to 4,000 apps, which sounds impressive until you ask whether your specific stack is on that list. Businesses running custom or legacy software will face gaps. Those gaps are often where the actual business value lives.
Third, trust and oversight become a management issue, not just a technology issue. Letting an AI agent take actions on your behalf requires knowing exactly what it can and cannot do, and having a way to audit what happened. OpenAI’s safety guardrails are a start. They are not a substitute for your own oversight practices.
AI agents are not science fiction anymore. They’re a subscription tier. The question isn’t whether they matter. It’s whether you’re ready to use them well.
Want to explore how AI agents could benefit your business? Let’s talk.