Particle.news

ChatGPT Voice Runs Multi-Step Workflows and Accesses Email, Calendar and Slack

The global mobile and web rollout links GPT‑Live audio to the GPT‑6 model family so users can delegate hands‑free tasks with admin review available for sensitive actions.

Overview

  • OpenAI rolled out the update on Wednesday, connecting real‑time GPT‑Live audio to ChatGPT’s mobile and web apps so users can trigger multi‑step background workflows by speaking.
  • Voice can now use plugins and connected apps such as Gmail, Google Calendar and Slack to scan inboxes, book meetings, summarize messages and perform other workplace tasks without manual input.
  • The voice layer is backed by the GPT‑6 suite (Astra, Sol and Luna), which the app selects based on task complexity to balance speed, cost and capability.
  • Feature access is tiered: Plus and Pro subscribers get Work‑tab voice tools to build docs, decks, sites and spreadsheets while Free and Go users retain plugin access, and Enterprise and Education accounts can require admin review for sensitive actions.
  • This move builds on July’s GPT‑Live desktop integration and signals OpenAI’s push toward hands‑free personal assistants and future device plans, but it raises trade‑offs between convenience, slower voice consumption and new privacy and security responsibilities.