As we increase our reliance on AI, decoupling intelligence from organization will become vitally important. The model providers are starting to build the systems through which agents communicate and coordinate; we could end up depending on them much more than we depend on the models.
In May 2025, Microsoft announced task delegation between agents in Copilot Studio and an Entra directory of agent identities. In July 2026, OpenAI announced a multi-agent beta in the Responses API, with hosted actions for spawning agents, sending messages, and waiting for replies. In August, Anthropic added [messaging between independent Claude Code
sessions](https://github.com/anthropics/claude-code/releases/tag/v2.1.224). These announcements point in the same direction. Agents that can communicate can divide work, question one another’s conclusions, and take on more responsibility without bringing every problem back to a human. It makes sense that the model providers want to build this into their products.
Consider what an organization invests in such a team. It defines roles, establishes who can authorize work, develops procedures, and accumulates knowledge. Agents learn where to find an answer and whom to ask when the documentation is insufficient. Tasks and conversations connect all of this. Teams learn, evolve and adapt.
These arrangements become the operating framework of the organization.
If they are coupled with the model provider’s infrastructure, changing model providers can mean rebuilding how the organization works. A better model elsewhere may not be enough to justify the move; a price increase, a withdrawn model, or a change in access conditions can impact the organization’s ability to operate. Agent identity is central here. A reviewer agent should remain the organization’s reviewer when we change the model or software running it, and retain its accumulated knowledge. Other agents should still know how to reach it, recognize its messages, and check what it is authorized to approve. Replacing the machinery should preserve those relationships.
The same applies to tasks, communication, and accumulated knowledge. They need to remain accessible, and under the organization’s control, across providers and harnesses. This should hold whether agents run through Gas Town, Gas City, or terminals we start ourselves.
Intelligence should become a commodity, and it should be decoupled from the infrastructure needed to make the most of it.
This is one reason I am building aweb. It provides identity and communication independently of the agents’ runtimes, along with a shared task system. It supports other arrangements, like keeping
tasks in [beads](https://github.com/gastownhall/beads) and using
[aweb for its mail](https://aweb.ai/docs/beads-mail/). Aweb also
has experimental support for A2A, an open protocol for exchanging tasks between agents built with different frameworks.
Open source lets an organization inspect, operate, and change the software it depends on. Open protocols let independently built components communicate. Neither removes the work of integration, but both give us ways to retain control as our needs change.
Can we replace a model, a harness, or a service and keep the team working? We should build toward that while the agentic organizations are still taking shape.
ps. Microsoft’s support for A2A is a welcome step in the right direction.