
Artificial intelligence is moving beyond simple prompt-and-response boxes. During its annual DevDay conference in San Francisco, OpenAI introduced Dots, a new fleet of always-on personal AI agents designed to handle daily work tasks, code debugging, meeting scheduling, and research entirely in the background.
Each dot operates on its own isolated cloud computer with a virtual browser powered by the company’s flagship GPT-6 Astra engine. Rather than waiting for live human prompts, these digital assistants run continuously, connecting to over 4,000 apps via plugins to manage multi-step projects around the clock.
Proactive research and strict approval controls
To keep agents helpful without turning intrusive, OpenAI designed Dots to learn user preferences through direct feedback. While working in the background—a mode OpenAI calls proactive research—a dot operates in read-only mode across connected apps, meaning it cannot send emails or alter external files without explicit authorization.
When an agent spots an actionable task, it prepares the work for human review. For instance, early tester Dan McAteer shared that his dot identified a forgotten freelance invoice, gathered thread details, drafted the document, and held it as a PDF until he gave approval to send it. Users can inspect their dot’s cloud desktop at any time, communicate via ChatGPT, Slack, or Teams, and establish custom rules regarding which actions require approval.

Subscription tiers, safety context, and enterprise rollouts
OpenAI is including one primary dot for ChatGPT Pro and Business Premium subscribers. Pro availability currently excludes the European Economic Area, the UK, and Switzerland, though (via The Next Web). Enterprise users can also test specialized dots built for roles in procurement, support, and contracting, with OpenAI partnering with Microsoft to bring these corporate bots into Agent 365.
The launch arrives alongside broader platform updates. That includes a sneak peek at GPT-6.1 Sol, a cheaper model that offers near-Astra intelligence at a fraction of the price of regular tokens. The launch of Dots also comes at the same time the industry is laser-focused on agent safety, just after OpenAI pulled the plug on its planned launch of GPT-6.1 Astra following internal testing over safety concerns.
The post OpenAI Unveils Dots: Always-On Personal AI Agents Powered by GPT-6 Astra appeared first on Android Headlines.
​Â