Jack Dorsey's Buzz: The Open-Source AI Chat Platform That Treats Agents as Colleagues

On July 21, 2026, Jack Dorsey's Block (formerly Square) launched Buzz — a free, open-source group chat and collaboration platform designed for teams of humans and AI agents . Unlike chatbots that wait for prompts, AI agents in Buzz operate as full team members with their own accounts, posting messages, reviewing code, and running workflows alongside people . Dorsey announced the platform on X, calling it "model-agnostic, decentralized, self-sovereign, and open source" .

What Buzz Actually Does

Buzz merges three tools that most development teams currently keep separate: a chat app (like Slack), a code repository (like GitHub), and an AI agent interface . Its interface looks familiar — channels on the left, a chat pane in the middle, a sidebar with context — but underneath it works differently .

Key product details:

• AI agents as full members: Agents sit in channels, post messages, review pull requests, and execute workflows as independent participants, not as add-ons or integrations . Buzz is agent- and model-agnostic, meaning it works with Anthropic's Claude Code, OpenAI's Codex, and Block's own Goose agent .
• Built-in code hosting and Git: Teams can reference code repositories, run automated workflows, and move from chat to planning to pull requests in one surface, reducing the need to jump between tools .
• Built on the Nostr protocol: Buzz runs on a Nostr relay, making it decentralized and self-hostable . This is a deliberate architectural choice that gives organizations control over their data and identity infrastructure, unlike centralized services such as Slack or Microsoft Teams .
• Cross-platform and free: Desktop apps are available for macOS, Windows, and Linux, and the source code is on GitHub at github.com/block/buzz . Organizations can self-host their own instance .

Block acknowledged that Buzz is "still in early stages" and recommended teams not fully migrate yet — the platform needs to prove stability, ecosystem appeal, and real collaboration efficiency .

How Buzz Fits Into Block's Broader AI Strategy

Buzz is not a standalone product launch. It is the most visible expression of a sweeping AI-first reorganization at Block that has been unfolding throughout 2026. Dorsey's stated philosophy runs through every piece of it: "We're going to build this company with intelligence at the core of everything we do" .

1. AI-first organizational restructuring

In February 2026, Block announced plans to cut approximately 40–50% of its workforce — roughly 4,000 jobs — as part of a strategic overhaul to integrate AI across all operations . Dorsey described the company reshaping itself into smaller, more focused teams powered by AI tools, arguing that "a much smaller team, utilizing the tools we are developing, can achieve more and perform better" . Block's stock surged over 16% on the announcement .

2. AI replacing middle management

In March 2026, Dorsey published a blog post co-authored with Sequoia Capital partner Roelof Botha titled "From Hierarchy to Intelligence" . The post argued that traditional corporate hierarchies — a two-thousand-year-old model originating from the Roman Army — exist primarily to route information, and that AI can now perform that coordination function more efficiently . Instead of humans relaying information through layers of management, Dorsey and Botha proposed that "a system can maintain a continuously updated model of an entire business and use it to coordinate work" . Their goal: a company built "as an intelligence (or mini-AGI)" .

3. Internal dogfooding

Dorsey said Block had been using Buzz internally for weeks before the public launch . The platform is a direct expression of his thesis: teams of humans and AI agents collaborating as equals in a decentralized, self-sovereign workspace, with agents accountable through signed events and shared identity .

4. Earlier open-source AI building blocks

Before Buzz, Block's Open Program Office released Goose in January 2025 — an open-source AI agent-building platform that lets developers integrate large language models into applications . By October 2025, Goose had been rolled out to all 12,000 Block employees within eight weeks, with engineers reporting time savings of 8 to 10 hours per week . Buzz functions as the collaboration-layer companion to Goose's agent-building layer, completing the stack.

5. Customer-facing AI products

Block has also embedded AI into its merchant and consumer products. Managerbot, a proactive AI agent for Square, reached roughly one million businesses by April 2026 . Money Boost (also called Money Bot), an AI-powered feature inside Cash App, drew one million active users in a single week with no marketing .

Buzz in the Emerging Market for AI-Native Workspaces

Buzz enters a market that is forming around a few converging themes:

• AI agents as colleagues, not tools: Unlike copilot-style assistants that only respond when prompted, Buzz treats agents as full workspace members with identities and accounts . This "agent-as-colleague" model is a distinct architectural bet compared to platforms that add AI as an integration layer on top of existing chat.
• Decentralized and self-sovereign: Buzz's Nostr-based design positions it against centralized platforms like Slack and Microsoft Teams, appealing to organizations that want to control their data, identity, and AI infrastructure .
• Open-source ecosystem play: By open-sourcing Buzz on GitHub, Block is trying to seed a community around its agent-collaboration paradigm, similar to how Slack and Discord built ecosystems through APIs and third-party integrations . Teams can modify the product to their needs, which is especially attractive to organizations that prioritize data control and self-hosting .
• Developer-first integration: Buzz merges chat, code hosting, and AI workflows into one surface, collapsing the stack of tools developers currently juggle — Slack for communication, GitHub for code, and separate AI chat tools for agent interaction . This is a direct response to fragmentation in the current developer toolchain .

The open question: will teams adopt it?

Buzz is brand new and unproven at scale. It competes with deeply entrenched incumbents — Slack, Microsoft Teams, Discord — and faces the classic chicken-and-egg problem of any new collaboration platform . Its differentiation — open-source, decentralized, and AI-agent-native — is also its adoption hurdle, since most teams today do not yet operate with AI agents as full workspace members.

The bet is a forward-looking one: as AI agents become more capable and common, the first workspace platform designed from the ground up for human-agent teams will have a structural advantage over platforms that bolt AI on as a feature. Buzz formalizes a workspace model that Block itself is betting its entire organizational structure on.