OpenAI releases dots, a persistent AI assistant integrated into ChatGPT Pro
The new assistant, available to Pro-tier subscribers, operates in the background to proactively manage tasks across calendars, emails, and computers.
1 source · cross-referenced
- OpenAI launched dots, a persistent AI assistant integrated into ChatGPT Pro tier subscriptions starting at $100/month.
- The assistant operates in the background, proactively managing tasks like calendar conflicts and deadlines, though timing and granular controls remain works in progress.
- dots is positioned as an evolution of OpenAI’s coding tools, repackaged to feel intuitive and always-on for users.
- Privacy controls for dots match ChatGPT’s existing settings, but granular permissions and data access remain limited.
OpenAI has integrated a new persistent AI assistant, called dots, into its ChatGPT Pro tier and higher subscriptions, with pricing starting at $100 per month. The assistant is designed to operate in the background, proactively managing tasks such as calendar conflicts, deadlines, and other routine work without requiring explicit prompts from users.
Alexander Embiricos, OpenAI’s enterprise product lead, described dots as part of a broader industry trend toward "convergent evolution" in AI assistants, where framing the interaction as a single, continuous conversation improves intuitiveness. The assistant is positioned as an evolution of OpenAI’s earlier coding tool, Codex, which required more manual configuration and oversight.
During a demonstration, Embiricos noted that the assistant’s proactive capabilities are still maturing. For example, it once alerted him to a scheduling conflict just three minutes before the event, highlighting the need for improved timing and reliability. He expects these issues to be addressed over time as the product develops.
The assistant’s functionality is enabled by connecting to users’ calendars, emails, computers, and other tools. Embiricos shared an anecdote about an engineer at a drone delivery startup whose dots assistant detected an outage, analyzed it, and provided actionable steps to resolve the issue before the engineer noticed the problem.
Privacy controls for dots align with ChatGPT’s existing settings, allowing users to disable model training and manage resource access. However, granular permissions remain limited; for instance, the assistant does not accept read-only access to cloud drives. Embiricos acknowledged feedback about overly conservative defaults and pushiness, noting that the team plans to refine these aspects over time.
The launch of dots follows a broader industry push into persistent AI assistants, with competitors like Meta’s Muse and startups such as Instinct also releasing similar offerings. Embiricos emphasized that OpenAI’s approach prioritizes making the technology feel intuitive and useful, even if it means repackaging existing capabilities rather than introducing entirely new technical breakthroughs.
- Nov 11, 2026 · Hugging Face
Hugging Face proposes token pooling-aware training to improve compressed multi-vector retrieval
Trust79 - Nov 11, 2026 · TechCrunch — AI
Fractile raises $720M Series C to develop AI chips and compiler stack
Trust79 - Nov 11, 2026 · The Verge — AI
Google’s Tabletop AI experiment lets users build custom collaborative apps via waitlist
Trust79