OpenAI launches workspace agents for business teams, may phase out GPTs

OpenAI released workspace agents for users on its Business, Enterprise, Edu, and Teachers plans on April 22, 2026. The cloud-based agents can perform business tasks autonomously, including finding product feedback on the web and sending reports in Slack, or drafting follow-up emails in Gmail.

The agents can be shared within organizations, allowing teams to build once and deploy across ChatGPT or Slack. According to OpenAI, the agents are designed to "gather context from the right systems, follow team processes, ask for approval when needed, and keep work moving across tools."

GPTs likely being phased out

OpenAI describes workspace agents as an "evolution" of GPTs—the custom chatbots it launched in 2023. The company stated that "GPTs will remain available while teams test workspace agents with their workflows" and will "make it easy to convert GPTs into workspace agents" soon. This language suggests GPTs may be discontinued after a transition period.

When asked for clarification, OpenAI spokesperson Taya Christianson directed to the blog post without providing additional information about GPTs' future.

Competition in agent space

The release follows increasing competition in AI agents. Anthropic offers Claude Cowork, an agent that completes tasks using files from users' computers, plus a separate platform for building autonomous agents.

OpenAI hired Peter Steinberger, founder of OpenClaw (formerly Clawdbot and Moltbot), an AI agent that went viral claiming to be "the AI that actually does things."

What this means

OpenAI is consolidating its product line by replacing GPTs with more capable workspace agents, betting that businesses want autonomous task execution over conversational custom bots. The move puts OpenAI in direct competition with Anthropic's agent offerings and signals a shift from chat interfaces to action-oriented AI tools. The phasing out of GPTs—less than three years after launch—shows how quickly OpenAI is willing to deprecate products when newer architectures emerge.