OpenAI is bringing voice-based agentic capabilities to the ChatGPT mobile app, extending work-oriented features that had previously reached its desktop app.
The update lets users initiate certain multi-step tasks by voice on a phone, including drafting documents and emails or summarizing Slack messages. It reflects a broader shift in AI interfaces: voice is moving beyond conversational questions toward directing software to complete work.
What is changing
For ChatGPT Plus and Pro subscribers, the mobile app’s Work tab will support voice-led workflows. OpenAI says those users can create documents, draft emails, summarize Slack messages, build websites and presentations, use a cloud browser, and access areas including finances through ChatGPT.
Free and Go users will be able to work with plugins and connected apps, according to the company.
OpenAI is also adding richer text output to voice conversations and making it easier to switch between voice and text. A user can begin a conversation while mobile and resume it on desktop, a handoff that matters more once a request becomes a document, research task or workflow rather than a short exchange.
Why it matters for teams
The practical value is not simply dictation. Mobile voice can reduce the friction of capturing and assigning work at the moment it arises: a manager could turn a meeting takeaway into a drafted follow-up, or a founder could begin assembling material for a presentation before returning to a computer.
That usefulness depends on the quality of the connected systems and the controls around them. Summarizing Slack messages, interacting with finance-related areas, or using a cloud browser can expose sensitive company information. Organizations considering these capabilities will need to examine which apps can connect, what data is available to the model, who can authorize actions, and how outputs are reviewed before they are sent or used.
The product packaging also makes the commercial boundary clear. OpenAI is reserving the Work-tab capabilities for Plus and Pro users, while offering free and Go users access to plugins and connected apps. Teams will need to determine whether the paid tier’s mobile workflow features deliver enough time savings to justify broader deployment.
Part of a wider interface contest
OpenAI introduced its GPT-Live conversational model in July and subsequently integrated it with the desktop app for tasks in the Work tab and app-building in Codex. This mobile rollout brings related functionality to the phone.
The competitive question is increasingly about continuity across devices and task types, not only model quality. Anthropic recently made handoffs between mobile and desktop easier and combined its Cowork and Chat interface. OpenAI, by contrast, continues to keep chat and workspaces separate.
That distinction may matter for adoption. A separate workspace can help contain work processes and permissions, but it can also add navigation and context-switching. A unified interface may be simpler, while potentially blurring casual chat and operational work.
What to watch next
The key test will be whether voice can reliably launch workflows without creating more review work than it saves. Watch for clearer details on permissions, auditability and approval steps for connected apps—especially where Slack, browser activity or financial information is involved.
Also watch how often mobile interactions successfully continue on desktop with their context and artifacts intact. That handoff, rather than the novelty of speaking to an assistant, is likely to determine whether voice becomes a durable work interface.
