At DevDay 2026, OpenAI turned the "always-on agent" rumor into a product: Dots — persistent agents with their own cloud computers that automate recurring work in the background. Here's the operator's reading: what's real, where they fail, and what deserves your attention.
Key takeaways
- Dots are always-on autonomous agents: each gets its own cloud computer and browser, works through 4,000+ connected apps, and keeps context over time.
- Standout detail: "proactive research" mode — background read-only access that can't send messages or change content. The right posture to start in.
- OpenAI's own warning: "Dots can still make mistakes, so always review consequential work." Approval gates aren't optional.
- Rollout: Pro and Business Premium users first (first Dot included free; Dot conversations don't count toward usage limits), Enterprise/Edu/Healthcare as admin-enabled beta.
- Bottom line: inventory your recurring work this week, write the brief before connecting tools, and instrument the baseline — measured agent work is an improvement; unmeasured is a demo.
Last week's column covered the rumor — an always-on OpenAI agent leaked ahead of DevDay. At DevDay 2026 on September 29, it became an announced product: Dots, always-on autonomous agents designed to automate recurring tasks and operate continuously in the background, learning user priorities and freeing users from one-prompt-at-a-time interactions. OpenAI unveiled 20+ updates across ChatGPT, Codex, and its model ecosystem; Dots led the announcements. (TechGig)
What follows is the operator's reading — what's real, where the failure modes are, and what deserves your attention.
What OpenAI actually shipped
A Dot is not another chat thread. OpenAI says each Dot gets its own cloud computer and browser, works through apps the user connects — more than 4,000 apps through its plugin ecosystem — and keeps context over time. You reach a Dot through ChatGPT on desktop, web, and mobile, and through Slack and Teams, with the same Dot experience across interfaces. (Digit.in)
The demonstrations focused on operational workflows, not consumer showcases: triaging recurring bug reports and feature requests, preparing pull requests, managing planning cycles, migrating an application off a retiring API. The product shifts the interaction model from one-off requests toward longer-running delegated tasks. (Substack / DevDay TLDR)
Two details that deserve your attention
Proactive research mode. In the background, a Dot can look through connected apps using read-only access to find information that may be useful — it cannot send messages or change content in this mode. The read-only posture is the right one to start in: narrower permissions for early experimentation.
Specialist dots for business. OpenAI is testing agents aimed at workflows involving company systems: procurement, invoice processing, customer support, commercial contracting. Those workflows are common automation candidates — repeatable processes, structured data — and they're where agent deployments get tested against reality. (Digit.in)
Availability and pricing
Dots are rolling out to Pro and Business Premium users in eligible markets, with Enterprise, Edu, and Healthcare access as a beta that a workspace administrator enables. The first Dot is included at no additional cost on Pro and Business Premium — and conversations with a Dot don't count toward ChatGPT usage limits. (Digit.in)
The detail that matters: your approval gates
OpenAI's own guidance is unusually blunt. Users can set custom rules, monitor progress, and decide which actions need approval — and certain sensitive actions (OpenAI names changing a password) will always require the user. And in OpenAI's own words: "Dots can still make mistakes, so always review consequential work." (Digit.in)
That statement is doing real work. A Dot with its own computer, browser, and thousands of connected apps is a background process with production access. Treat it like a production service: named owner, approved action list, reviewed logs.
Where always-on agents fail
The failure modes of background agents are specific and worth naming:
- Unwatched wrong actions — a mistake executed while nobody's looking
- Ambiguous instructions treated as rules — vague briefs interpreted literally
- Tool-call loops — burning usage on circular work
- Silent failures — a task quietly stops completing
- Stale context — decisions made on outdated information
That list dictates your selection criteria. Good first candidates: repetitive, measurable, reversible, low-risk — report generation, triage, follow-ups. Poor candidates: ambiguous decisions, regulatory judgments, irreversible actions.
Before deployment, define agent ownership, access boundaries, and review procedures — and settle who owns the brief when the person who wrote it changes roles.
The pricing model lowers the barrier for experimentation
The first Dot ships included with Pro and Business Premium at no additional cost, and Dot conversations don't consume usage limits. That's a wedge: embed always-on agents in recurring workflows first, price the fleet later.
But the subscription is the cheapest line on the bill. Integration engineering, data cleanup, workflow redesign, security review, and monitoring will dwarf it. Budget the deployment, not just the seat.
DevDay's other announcements point the same way. GPT-6.1 Sol targets near-Astra intelligence at one-fifth of Astra's token prices. A new $500/month Pro 500 tier offers 25x the Plus allowance with Astra Ultrafast — 300 tokens per second in Codex. (Analytics Insight) The trend: lower-cost access to capable models, with latency as the premium feature. Keep the integration layer vendor-neutral — competitors will follow.
What this means for your business
Inventory your recurring work this week. Bug-report triage, report generation, follow-ups, invoice chasing, status compilation — list the tasks that stall waiting for someone's attention. No new product is needed to make the list.
Write the brief before connecting the tools. Objective, allowed actions, forbidden actions, required approvals, success criteria. Start with read-only access — proactive research mode is effectively that posture — until the agent has earned write access.
Name the operating model. Who reviews the outputs? Who maintains the agent's instructions when processes change? Who handles exceptions, and who audits the logs? An agent that changes tasks also changes roles — assign them.
Instrument the baseline now. Cycle time, hours spent, error rates, escalation frequency — measure before the agent touches anything. Agent work that isn't measured against a baseline is a demo, not an improvement.
The bottom line
Always-on agent workflows are moving from prototypes into commercial products. The organizations that benefit most will be the ones that defined clear workflows, permissions, and success metrics before the agent got its keys.
Want a grounded map of which of your workflows an always-on agent can actually own — and which ones it can't? Get a free AI audit and we'll work it through with you.
FAQ
What are OpenAI Dots?
Always-on autonomous agents announced at DevDay 2026 (Sept 29). Each Dot gets its own cloud computer and browser, connects to 4,000+ apps, keeps context over time, and automates recurring tasks in the background — reachable via ChatGPT, Slack, and Teams.
How much do Dots cost?
The first Dot is included free on Pro and Business Premium plans, and Dot conversations don't count toward ChatGPT usage limits. Enterprise, Edu, and Healthcare get beta access enabled by a workspace admin. The real cost is deployment: integration, data cleanup, workflow redesign, security review, monitoring.
What is "proactive research" mode?
A background mode where a Dot looks through connected apps with read-only access to find useful information — it cannot send messages or change content. It's the recommended starting posture: narrower permissions while you evaluate.
What are the main failure modes?
Unwatched wrong actions, ambiguous instructions interpreted as rules, tool-call loops burning usage, silent failures, and stale context. Mitigation: start with repetitive, measurable, reversible tasks; define ownership, access boundaries, and review procedures before deployment.
Sources
Sources
- OpenAI DevDay 2026 brings new AI agents, developer toolstechgig.com
- OpenAI dots announced at DevDay 2026: How these always-on AI agents will handle tasks on your behalfdigit.in
- OpenAI DevDay 2026: 20+ AI Tools, GPT-6.1 Sol, Dotsanalyticsinsight.net
- OpenAI DevDay Keynote TLDRsifuyik.substack.com
Want to know what this means for your stack? A free AI audit maps your workflows and shows where automation pays off — in your numbers, not ours.