Skip to content
Sign in
Technology · North America

OpenAI launches dots assistant for higher-tier ChatGPT subscribers

Higher-tier ChatGPT subscribers can now use OpenAI’s dots assistant, as the company’s assistant lineup evolves; OpenAI was scheduled to stop accepting new user-created Custom GPTs on October 26, 2026.

AI-assisted synthesis · Based on 4 outlets' reporting · Article text read from 3 of 4 outletsHow it's made

View timeline · 10 entries
4 outlets2 regionsLatest reporting 4 min read
Coverage: Center OnlyL 0 · C 4 · R 0
File photo
Photo: Coolcaesar / Wikimedia Commons (CC BY 4.0)

What happened

Key facts

  • OpenAI released dots, its assistant offering, at its developer day in late September (per TechCrunch).
  • Dots is described as an always-on AI assistant for paid ChatGPT users (per TechCrunch, tradersunion.com).
  • Dots is available through the ChatGPT desktop and mobile apps to subscribers on the Pro tier or higher (per TechCrunch, pluang.com, techtimes.com).
  • OpenAI product executive Embiricos said dots includes a feature the company calls “proactive research.” (per TechCrunch).
  • Embiricos said OpenAI needs to develop highly granular controls for dots (per TechCrunch, pluang.com, techtimes.com).
  • TechTimes says OpenAI was scheduled to stop accepting new Custom GPTs on October 26 (per techtimes.com, pluang.com).

The brief

OpenAI has launched dots, an assistant offering for ChatGPT subscribers on the Pro tier and above. OpenAI product executive Embiricos described the tool as part of a wider push to build AI assistants, while the service is available through ChatGPT’s desktop and mobile apps. TechCrunch reported that OpenAI released dots at its developer day in late September.

TechTimes gives a more precise account, saying OpenAI announced Dots on September 29, 2026, during its annual DevDay conference at Fort Mason in San Francisco. TechCrunch’s excerpt focuses on the assistant’s paid availability and product features; TechTimes places the announcement among a series of other company announcements.

The subscription threshold is one of the clearest practical limits: TechCrunch says access begins with the Pro tier or higher, with plans starting at $100 a month. Traders Union’s headline also characterizes dots as an always-on assistant for paid ChatGPT users.

The Indonesian-language Pluang excerpt identifies dots as an AI assistant intended to help users manage tasks and schedules, but gives no subscription or product-feature details. Embiricos said dots includes “proactive research.” He also said the product team needs to build highly granular controls, describing control design as an important product issue.

His comments frame the launch not just as an assistant feature but as a question of how users can manage the tool’s operation. The accounts offer different levels of detail rather than conflicting launch dates: TechCrunch places the release in late September, while TechTimes specifies September 29.

TechTimes additionally reports that OpenAI was due to stop accepting new Custom GPTs on October 26, describing those as specialized chatbots created by developers. It does not explain how that change relates to dots, so the two developments should be treated as separate company actions.

The sources do not report a response from ChatGPT subscribers or describe the number of dots users, nor do they establish how the planned Custom GPT change affects existing bots. The immediate questions raised by the reporting concern who can access dots under the paid-tier structure and how OpenAI will implement the granular controls Embiricos says the product needs.

Why it matters

  • Dots could shift ChatGPT from a tool people consult on demand to an assistant that monitors ongoing tasks and surfaces research proactively, potentially saving time for professionals who can delegate routine information-gathering. Because access is limited to Pro-tier subscribers and above, those benefits initially accrue to higher-paying customers, while OpenAI tests whether persistent assistance justifies premium subscriptions. The need for granular controls also makes users’ ability to limit what the assistant monitors a key condition for adoption.

What to watch next

  • Whether OpenAI provides the highly granular controls Embiricos said dots needs.
  • How OpenAI presents dots to subscribers on the Pro tier and higher through the desktop and mobile apps.
  • Whether OpenAI stops accepting new Custom GPTs on October 26, as TechTimes reported.
  • Whether OpenAI explains how the Custom GPT change relates to dots.

How it's being covered

Coverage analysisShow details
4 source articles · 4 unique outlets
Outlet political mix: Center Only· 4 Center
Where the outlets are based: US/Canada 50% · Asia 25% · Other 25%
Outlet political mix
Center Only
Left
Center
Right
0 outlets
4 outlets
0 outlets
The marker shows coverage balance: each unique outlet counts once.
Geography
Where the outlets covering this story are based.
4 unique outletsDominant coverage: US/Canada
Coverage imbalance
  • Coverage is currently concentrated among Center-rated outlets (4 of 4 rated sources); comparatively few Left or Right-rated outlets have published on the event.
How we measure coverage

How outlets' accounts compare

2 framing differences · 2 uncorroborated claims · 2 information gaps
The reports repeat the same account with different emphasis; its central claims are not independently confirmed in the reporting Updat3 read.Compared across 3 reports whose article text Updat3 read; 1 other outlet was checked by headline only.
Framing / emphasis differences
Sources emphasise different angles, actors or consequences of the same reporting.
  • TechCrunch emphasizes paid access, the mobile and desktop apps, proactive research, and the need for detailed user controls; Pluang describes dots as helping users manage tasks and schedules.
  • TechTimes gives the announcement date and venue and situates Dots among other product announcements; TechCrunch describes the release as occurring in late September.
Updat3: These are differences of emphasis, not disputes about what is reported. They do not show that the reported claims are true.
One-sided or uncorroborated claims
Claims that come from one party or one source family and are not independently corroborated.
  • Embiricos said OpenAI calls one dots feature “proactive research.”
  • Embiricos said OpenAI needs to develop highly granular controls for dots.
Updat3: Not independently corroborated yet — treat as a claim, not an established fact.
Not established by the reporting
Information the available reporting does not contain or confirm. This is a gap, not a disagreement.
  • The sources do not quantify dots adoption or describe user outcomes.
  • The sources do not establish how the planned change to new Custom GPT submissions relates to dots.
Updat3: Updat3 cannot conclude anything about this from the current sources.

Timeline

Developments from this story's sources, dated background we compiled and fact-checked, and our earlier coverage. Each row says which.
  1. OpenAI was founded as an artificial intelligence research organization.
    Background · compiled and fact-checked by Updat3
  2. ChatGPT became publicly available on November 30, 2022, using a conversational format that let people ask follow-up questions and refine requests.
    Background · from our historical context
  3. OpenAI released ChatGPT to the public.
    Background · compiled and fact-checked by Updat3
  4. In February 2023, it introduced ChatGPT Plus, a paid plan that offered benefits such as access during busy periods and earlier availability of new features.
    Background · from our historical context
  5. In August 2023, it launched ChatGPT Enterprise for organizations seeking additional security, administration, and higher usage limits.
    Background · from our historical context
  6. At its first DevDay, on November 6, 2023, OpenAI introduced “GPTs”—custom versions of ChatGPT that users could configure for particular purposes, including by supplying instructions and reference material.
    Background · from our historical context
  7. OpenAI introduced ChatGPT memory in 2024, enabling the service to retain certain useful details across conversations, with controls for reviewing or deleting what it remembered.
    Background · from our historical context
  8. The GPT Store, which opened in January 2024, gave users a way to discover and share some of these specialized assistants.
    Background · from our historical context
  9. OpenAI launched ChatGPT Pro in December 2024 at a higher price than Plus, targeting users who needed more extensive access to advanced models and tools.
    Background · from our historical context
  10. OpenAI introduced ChatGPT agent.
    Background · compiled and fact-checked by Updat3

How we got here

OpenAI’s move toward assistants that can remain available beyond an individual chat grew out of a product that initially centered on answering prompts in a text window. ChatGPT became publicly available on November 30, 2022, using a conversational format that let people ask follow-up questions and refine requests. Its rapid adoption helped establish chatbots as general-purpose tools, but it also exposed a limitation: users often had to restate their goals, provide background, and initiate each interaction themselves. OpenAI began adding ways to make ChatGPT more tailored and continuous. In February 2023, it introduced ChatGPT Plus, a paid plan that offered benefits such as access during busy periods and earlier availability of new features. In August 2023, it launched ChatGPT Enterprise for organizations seeking additional security, administration, and higher usage limits. At its first DevDay, on November 6, 2023, OpenAI introduced “GPTs”—custom versions of ChatGPT that users could configure for particular purposes, including by supplying instructions and reference material. The GPT Store, which opened in January 2024, gave users a way to discover and share some of these specialized assistants. These developments broadened the idea of an assistant from one general chatbot toward tools adapted to different tasks and users. Other additions addressed the friction of using a chat interface. OpenAI introduced ChatGPT memory in 2024, enabling the service to retain certain useful details across conversations, with controls for reviewing or deleting what it remembered. It also developed voice and image capabilities, including Advanced Voice Mode, which became broadly available to paid users in September 2024. Such features made interaction less dependent on typing and a single exchange, while raising questions about how much context an AI system should retain and when it should act on that context. The subscription structure also became more closely tied to access to advanced capabilities. OpenAI launched ChatGPT Pro in December 2024 at a higher price than Plus, targeting users who needed more extensive access to advanced models and tools. Later product additions increasingly combined model access with functions such as searching, handling files, and carrying out multistep work. In this setting, an always-available assistant represents a further change in how the service may be used: rather than only responding when a user opens a chat and asks a question, it can be positioned to help anticipate or develop research needs. OpenAI product executive Embiricos described one such capability as “proactive research.”

That shift makes user control a central product issue, not merely a technical detail. An assistant that works across desktop and mobile apps and can initiate useful work must distinguish between helpful anticipation and unwanted interruption, and users need clarity about what information it can use and what actions it may take. Embiricos has said OpenAI needs to develop highly granular controls for the service. The emphasis on those controls reflects a broader challenge for AI companies: making assistants more persistent and capable while giving users meaningful choices over their behavior.

Original sources

4 articles from 4 outlets

Every story links to the reporting it was built from. How we choose sources

Spotted a factual error, a broken link or a misleading summary? Tell us.