#

Key takeaways

• OpenAI announced dots on September 29, 2026.

• Dots are persistent agents powered by GPT-6 Astra.

• They use their own cloud computers and connected applications.

• Users can set permissions, approval rules, and blocks.

• The source does not confirm Moroccan availability.

What OpenAI announced

OpenAI announced dots on September 29, 2026. The company describes them as persistent agents powered by GPT-6 Astra. They work using their own cloud computers and connected applications.

The launch page says dots can continue tasks over time. They can also learn from user feedback. OpenAI presents them as always-on agents that keep working across sessions.

How dots are meant to work

OpenAI says its plugin ecosystem connects to more than 4,000 applications. Users can interact with dots through ChatGPT, Slack, and Teams. They can also inspect the agent's computer while work proceeds.

The source gives examples involving development, analysis, and content preparation. These examples show intended workflows. They do not prove the same results for every customer.

Controls, permissions, and safety

OpenAI says users can choose which connected applications dots may access. Custom rules can permit actions, require approval, or block them. Built-in safety requirements still apply.

The company also says security monitoring can pause or stop work when concerns arise. Dots can connect to a user's laptop only with permission. That suggests the product is designed around user control rather than open-ended access.

Background proactive research

OpenAI distinguishes background proactive research from other activity. In that mode, connected tools are restricted to reading. The research cannot send messages or change application content.

That limit matters because it narrows what the agent can do on its own. It also shows that OpenAI is separating observation from action. The source presents this as a safety boundary.

Specialist dots for organizations

OpenAI also previews specialist dots for organizations. These have dedicated identities and deeper integration into business systems. The source does not give more detail on how those identities work.

This part of the announcement suggests a broader enterprise direction. It points to agents that can be tailored for organizational use. The source does not confirm any specific deployment model beyond that.

Availability and rollout

The launch page says rollout begins across Pro, Business Premium, and Enterprise plans in eligible markets. That wording does not establish access in every country. The source does not announce Moroccan availability.

It also does not mention a local partnership, local hosting, or compliance with any particular Moroccan rule. Readers should treat access details as limited to the source's wording. Any broader availability would need separate confirmation.

Morocco relevance

The source reports none. For readers, the global lesson is simple: always check eligibility, permissions, and tool access before adopting an always-on agent.

Why this launch matters

Dots combine persistence, connected apps, and user controls in one product. That mix makes them different from a one-off chatbot session. The agent can keep working while the user reviews progress.

The announcement also shows how OpenAI is framing automation. It emphasizes permissions, monitoring, and restricted research modes. Those details matter because they shape how much autonomy the agent actually has.