ChatGPT mobile app gets voice-based agentic features
TechCrunch Ivan Mehta
OpenAI put voice-powered work tools into ChatGPT on phones. Now people can start tasks there and keep them going on desktop.
Based on reporting by TechCrunch, Ivan Mehta — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
OpenAI said Wednesday it is bringing voice-based agentic features to ChatGPT on mobile, turning the app into more than a place to chat back and forth. The new setup lets users trigger workflows such as drafting documents, writing emails, and summarizing messages by voice.
For Plus and Pro subscribers, the Work tab on a phone can be used to create a document, draft an email, or summarize Slack messages. Those users can also build sites, make presentations, use the cloud browser, and reach other areas in ChatGPT, including finances. Free and Go users get a narrower version, centered on plugins and connected apps.
OpenAI is also making voice responses feel less like a dead end. The company said voice conversations will now produce richer text output, and Plus users can switch between typing and speaking more easily. A task started on the move can now be resumed on desktop, which is the kind of small convenience that becomes the whole product once people get used to it.
The timing tracks a broader shift: more people are using voice to run complex commands through AI assistants. OpenAI’s earlier GPT-Live rollout in July brought voice into the desktop app, where it could be used in the Work tab or the Codex tab. This mobile update carries that same idea over, although OpenAI is still keeping chat and workspaces separate. Anthropic has been moving in the same direction with easier handoff between mobile and desktop and a merged Cowork and Chat interface. OpenAI, at least for now, is taking the more divided route.
My take — AI-written commentary, not fact-checked reporting
This is the sensible kind of AI feature: useful, slightly boring, and likely to stick. Voice works best when it gets people from request to result without making them babysit the interface. The real test is whether the separation between chat and workspaces feels neat or just like one more menu to learn.
Read more about this at: TechCrunch