Key Takeaways

  • Block’s Buzz gives humans and AI agents a shared workspace with cryptographic identities and auditable activity.
  • The open-source, self-hosted architecture targets enterprises seeking greater control over AI workflows and sensitive data.
  • Buzz enters a crowded market led by Slack and Microsoft Teams, while also competing with Mattermost and Rocket.Chat.

Block has launched Buzz, an open-source collaboration workspace designed to let employees and AI agents operate with similar capabilities inside a shared environment. Built on the decentralized Nostr protocol, Buzz assigns cryptographic identities to both people and bots, creating a verifiable record of who performed each action.

The launch arrives shortly after Block cut more than 4,000 jobs due to AI, giving the product an uncomfortable but important backdrop. Block is not merely introducing another chatbot interface. It is proposing a working environment in which software agents can participate more like colleagues, including opening repositories, submitting patches, reviewing code, running workflows, editing canvases, joining voice huddles, creating channels, and coordinating other agents.

That scope puts Buzz in competition with more than Slack. Block also sees the workspace as a possible alternative to GitHub, workflow dashboards, and the assorted integration layers connecting development and communication systems.

According to The Register, Block describes Buzz as free, open source, self-hostable, and more auditable than conventional chat products. In this context, Block uses “sovereign” to describe an environment that customers can host and control rather than leaving data, identities, and agent permissions entirely inside a proprietary service.

Enterprise buyers increasingly prioritize where AI agents run, which systems they can access, and how their actions are reconstructed after the fact. Those questions become particularly sensitive in fintech, healthcare, government, and other regulated sectors. A bot that summarizes a meeting presents one level of risk. An agent that modifies code, launches a workflow, or recruits another agent introduces elevated operational and security risks.

Buzz’s cryptographic key pairs could help enterprises distinguish human activity from automated activity while preserving a common audit trail. Related identity concepts are also appearing through OWASP’s Agent Name Service, DNS for AI Discovery, and Estonia’s work on digital IDs for agents. Block’s contribution is to place that identity model inside a broader collaboration environment.

“Every company is going to need a place where humans and agents work together,” the head of AI capabilities at Block said in a statement. “The question is whether that place is proprietary or open. We built Buzz because we believe it should be open.”

Major collaboration platforms are building parallel capabilities. Slack Developer Docs outlines agent experiences that operate inside Slack while using its interface, context, and permissions. Microsoft Teams is similarly evolving from messaging and meetings toward AI-assisted work. Mattermost and Rocket.Chat, meanwhile, appeal to customers seeking on-premise or self-managed collaboration environments.

Research projects are probing the same territory. Salesforce AI Research has published SlackAgents, a framework exploring agents that interact through Slack-like environments. Together, those efforts illustrate why Buzz is less a speculative oddity than an entry into a forming product category: the AI-native collaboration workspace.

Still, putting humans and agents in the same channels creates practical problems. Bots can generate messages, code changes, and task updates far faster than employees can review them. Without rate limits, approval gates, and carefully scoped permissions, a collaborative channel could become noisy or operationally risky. Who supervises an agent that orchestrates several other agents? And when should work pause for human review?

Block’s broader bet is that one shared substrate can replace the fragmented mix of chat, code forges, search indexes, continuous integration dashboards, and release systems used by many technical teams. Consolidation may improve context and auditing, but enterprises will likely evaluate whether Buzz can match mature incumbents on administration, integrations, reliability, and compliance controls.

Open source gives Block a credible point of differentiation, particularly for customers wary of vendor lock-in or restrictions around third-party AI services. It does not remove the adoption challenge. Slack, Microsoft Teams, Mattermost, and Rocket.Chat already have installed bases and expanding agent features. Buzz now has to show that cryptographic identity and self-hosting produce a workspace that is not only more controllable, but also usable when humans and machines start working at very different speeds.