OpenAI expands dots as always-on AI assistant for paid ChatGPT users

Traders Union
All News
Business & Wealth
Companies
OpenAI expands dots AI assistant

Highlights

OpenAI launches 'dots', a proactive always-on AI assistant for ChatGPT Pro subscribers with plans starting at $100 a month, aiming for broader task integration.
Current privacy controls for dots mirror ChatGPT's but lack granularity, with OpenAI planning finer controls after feedback and balancing user discretion.
OpenAI is actively engaging with payment processors to enable business adoption of AI agents, raising concerns over consumer authority and financial access.

Product strategy and rollout

As reported by TechCrunch, OpenAI released dots at its developer day in late September and is making the assistant available through the ChatGPT desktop and mobile apps for subscribers on the Pro tier or above, with plans starting at $100 a month.

Alexander Embiricos, who leads OpenAI’s enterprise product team, says the product is designed to present AI as a single persistent entity rather than a series of separate chats. He says that framing, along with the mascot-style presentation and branding linked to the negative space in OpenAI’s logo, is meant to make the technology feel more intuitive and eventually ubiquitous.

Embiricos says much of dots’ functionality can already be approximated through OpenAI’s Codex coding tool, but doing so requires more manual work. The company is repackaging capabilities built for software engineers into a more accessible assistant that can connect to calendars, email, computers and other tools to help users complete tasks with less supervision.

The proactive model is central to the offering. Embiricos says the assistant can review what it knows about a user, assess the resources it has been allowed to access and conduct what OpenAI calls “proactive research” to suggest ways it can be more useful, a design that also supports higher token usage for OpenAI’s platform.

Privacy and scaling concerns

OpenAI says dots uses the same privacy controls as ChatGPT, allowing users to disable model training and manage what resources the assistant can access on their computers. Even so, the current controls are not highly granular, and the product can be forceful in its permission requests, including by refusing read-only access to cloud drives.

Embiricos says OpenAI needs to build much more granular controls over time, while balancing that with an assistant capable of acting with discretion. He says the company launched with conservative defaults that asked for permission frequently, expecting to adjust those settings as it gathers feedback.

Performance also remains uneven. In testing described in the report, dots can identify calendar conflicts and deadlines, but not always with enough notice, although Embiricos says he expects the timing to improve. He also points to an example from a drone delivery startup where an assistant allegedly detected, analyzed and surfaced information about an outage before a human noticed it.

Beyond product design, OpenAI is also confronting the broader question of whether AI agents can scale across businesses, websites and payment networks. Embiricos says the company is working closely with the wider ecosystem rather than moving fastest on every new agent capability, and adds that OpenAI is in discussions with payment processors, a step that could expand commercial use cases but may intensify consumer concerns over how much authority and financial access such assistants should receive.

Our earlier coverage on enterprise AI agents highlighted that large U.S. companies are seeing mostly incremental benefits so far—efficiency and growth gains in targeted areas—while only a small share of CEOs currently view agentic systems as truly transformational. We also noted that adoption is being paired with continued investment in AI infrastructure and talent, as well as retraining and process changes needed to scale agents safely and effectively across core workflows.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.