Key points
- OpenAI released dots, an always-on personal assistant, at its developer day in late September, built from its Codex coding tools.
- It is available in ChatGPT desktop and mobile apps for Pro tier subscribers or higher, with plans starting at $100 a month.
- Once connected to calendar, email and computer, it runs in the background and proposes help through a feature called proactive research.
- OpenAI's Alexander Embiricos said the company is in talks with payment processors; privacy controls match ChatGPT's and are not yet granular.
Why it matters for marketers: Persistent assistants that read email and calendars will shape how work gets done and how data is collected, so marketers should watch how agents handle tasks, permissions and payments.
OpenAI has launched dots, an always-on personal assistant that keeps working without being prompted. According to TechCrunch, it was introduced at the company's developer day in late September and is available in the ChatGPT desktop and mobile apps to Pro subscribers or higher. Once connected to a user's calendar, email, computer and other tools, it tries to move work forward on its own.
Alexander Embiricos, who leads OpenAI's enterprise product team, told TechCrunch Disrupt that dots grew out of the company's agentic coding tools, and that much of its function was already possible with the Codex app, but with more hand-holding. A feature called proactive research periodically reviews what the assistant knows about a user and proposes ways to help. TechCrunch's testing found the foresight often useful but not always timely, such as late calendar-conflict alerts.
For marketers, the shift matters because assistants like this sit inside inboxes, calendars and files, which is where campaigns, approvals and client work live. TechCrunch notes the tension between the personal data an agent needs and what a software company collects, and points to criticism of Meta's Muse. OpenAI says dots has the same privacy controls as ChatGPT, including turning off model training, though the reporter found them not very granular and the app unwilling to accept read-only access to cloud drives.
Watch two things next. Embiricos said OpenAI wants far more granular controls. He also said the company is in conversations with payment processors, which could eventually let assistants spend money. Teams testing agents should review what access they grant and how clients and vendors feel about dealing with them.
Sources
This brief is our own summary. Read the full reporting at the source.
