myynti@metsosivut.fi

"*" indicates required fields

041 498 9805
Lets start conversation ➤
open ai dots

OpenAI Dots – What Does Dots Mean in Practice?

OpenAI Brings AI Agents Closer to Everyday Users.

OpenAI introduced a new Dots feature to ChatGPT at the end of September 2026.

A Dot is an AI assistant designed for continuous use and powered by the GPT-6 Astra model. Instead of being given only individual tasks, it can also be assigned longer-term responsibilities.

With regular ChatGPT, you would usually say:

“Do this.”

With a Dot, you can say:

“Take care of this from now on.”

What Is a Dot in Practice?

A Dot has its own cloud-based computer and browser. It can use connected applications, conduct research, perform scheduled tasks and make use of memory built from previous work.

For example, a Dot can:

Many of these capabilities are already familiar from existing agents. The main difference with Dots is continuity: they are not intended only to complete a single task, but to manage the same area of responsibility over a longer period of time.

A Dot Can Also Act Proactively

Dots includes a feature called proactive research.

It allows a Dot to review information from connected services and identify things it could help with, even when the user has not just given it a new task.

Proactive research running in the background uses read-only permissions. This means a Dot can review information, but it cannot use this feature to send messages or modify content in connected services.

This is an important safety boundary when an agent has access to services such as email, calendars or company systems.

What Is a Dot Allowed to Do?

Rules can be defined for a Dot to determine what it is allowed to do independently and when it must ask for permission.

For example, the system can be configured so that a Dot:

For example, reading a calendar can be allowed automatically, while sending an email to a customer can require approval.

OpenAI also points out that a Dot can make mistakes. Its work should be reviewed, especially when customers, money, contracts or other business-critical matters are involved.

OpenAI Is Making Agents Easier to Use

Perhaps the most interesting change introduced by Dots is not technical.

AI agents existed before Dots, but using them could easily feel technical and complicated. People talk about agents, tools, permissions, workflows and automations.

Dots aims to make all of this much simpler for ordinary users.

Instead of requiring users to build an agent themselves or understand how the underlying system works, they can simply tell a Dot in ordinary language what they want it to do.

For example:

“Check my calendar every morning and tell me the most important things for the day.”

Or:

“Monitor my household’s recurring expenses and tell me if there are areas where I could save money.”

A Dot is given a name and a persona, the necessary applications are connected to it, and its tasks are defined using ordinary language.

This may be one of the most significant aspects of Dots:

OpenAI is trying to make AI agents easy to use even for people who have never used agents before.

What Kinds of Dots Are There?

In addition to personal Dots, OpenAI is developing agents specialized for specific types of work.

Personal Dot

A Personal Dot is the user’s own personal agent.

It can be given a name, a persona, permissions and ongoing tasks.

For example:

“Monitor the progress of this project and notify me if anything important changes.”

Or:

“Check the open tasks every Friday and tell me what looks likely to carry over into next week.”

OpenAI has also discussed the possibility of using multiple Dots side by side in the future. In this model, different agents could be responsible for different tasks instead of relying on one general-purpose agent.

Specialist Dots

For businesses, OpenAI is developing Specialist Dots intended for specific job functions and business processes.

However, they are still in preview and pilot stages rather than finished products that any company can immediately adopt.

Examples include:

Accounting Dot

Accounting Dot is designed for accounting and financial administration tasks such as invoice processing, reconciliations, data verification and preparation for reporting.

The goal is not necessarily to replace an accountant, but to automate repetitive and time-consuming stages of the work.

Legal Dot

Legal Dot is designed for legal tasks and for helping ensure compliance with laws and regulations.

It can assist with tasks such as handling contracts, reviewing documents and comparing terms.

Marketing Dot

Marketing Dot is designed for marketing tasks.

Possible use cases include email marketing, campaign preparation, content processing and gathering marketing data.

What Other Kinds of Dots Could There Be?

OpenAI has also used software development, research, sales, customer service, procurement and contract processes as examples of potential Dots use cases.

In the future, companies could therefore have specialized agents for different functions, such as:

However, not all of these have been released as separate products. At this stage, it is more accurate to describe them as possible roles and use cases rather than as a finished catalog of Dot products.

Teams of Multiple Dots in the Future

OpenAI has also discussed a model in which multiple Dot agents work together.

For example, a company could have one agent for finance, another for marketing and a third for legal tasks.

Instead of one general-purpose agent, the company would then use a team made up of several specialized agents.

However, this is still a direction for the future rather than a finished operating model.

Dot vs. ChatGPT Work

ChatGPT can already perform multi-step agentic tasks.

ChatGPT Work can use files and applications, conduct research and produce outputs such as reports, documents and spreadsheets.

The main difference is how they are used.

ChatGPT Work

“Do this project.”

Dot

“Take care of this from now on.”

The distinction is not completely clear-cut, because Work can also handle longer-term tasks.

Dots, however, place greater emphasis on continuity, memory, scheduled tasks and ongoing areas of responsibility.

A Dot can also use ChatGPT Work and Codex as part of its work. In practice, a Dot can therefore act as a longer-term agent that makes use of other OpenAI tools when needed.

Dot vs. Grok Bot

OpenAI is not the first company to build an AI agent designed to operate continuously.

xAI launched Grok Bot in August 2026, and its basic idea is very similar.

Grok Bot has its own computer, can use different applications, can continue working without constant user guidance and can make use of previous context.

Dots and Grok Bot represent the same broader trend: the use of AI is shifting from writing individual prompts toward delegating tasks and responsibilities.

Does OpenAI Dots Work in Finland?

There is currently one significant limitation for Finnish users.

At launch, Dots is not available to ChatGPT Pro users in the European Economic Area.

This means that a Finnish ChatGPT Pro user cannot currently access a personal Dot.

Availability is broader for business use.

Business Premium users can access Dots in supported regions, while Specialist Dots are being piloted separately with companies.

What Does This Mean in Practice?

Dots is part of a broader shift in which AI agents are moving from individual tasks toward ongoing areas of responsibility.

xAI’s Grok Bot and other similar services are moving in the same direction. Competition is likely to accelerate development in usability, capabilities and safety.