OpenAI has decided that its Business, Enterprise, Edu, and Teachers plan users might actually want their AI to, you know, do things. So now they can create cloud-based “workspace” agents in ChatGPT that perform business tasks. Examples from OpenAI’s blog post include an agent that scours the web for product feedback and fires off a report in Slack, and a sales agent that drafts follow-up emails in Gmail - because apparently, we’ve all been spending too much time doing those tasks ourselves.
This move comes amid a surge of interest in AI agents, especially after OpenClaw - the AI agent formerly known as Clawdbot and Moltbot that boasts it’s “the AI that actually does things” - went viral. OpenClaw founder Peter Steinberger now works for OpenAI, which is convenient. Meanwhile, OpenAI faces competition from Anthropic, which offers its own Claude Cowork agent that can complete tasks using your computer files, plus a separate platform for making autonomous agents. Apparently, the AI arms race is now about who can make the most useful digital minion.
These new workspace agents can be shared within organizations, so “teams can build an agent once, use it together in ChatGPT or Slack, and improve it over time.” OpenAI assures that 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.” In other words, they’re like interns that never complain about coffee runs.
However, this development might spell doom for OpenAI’s “GPTs,” the custom chatbots announced in 2023. The company says workspace agents are an “evolution” of GPTs and that “GPTs will remain available while teams test workspace agents with their workflows.” Soon, OpenAI will also “make it easy to convert GPTs into workspace agents.” When asked for comment, spokesperson Taya Christianson pointed back to the blog post and offered no further insights - presumably because she was busy testing a workspace agent.