OpenAI wants its new tool to do your work for you and with you

· Source: AI - Ars Technica · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems · Depth: Intermediate, short

Summary

OpenAI has launched ChatGPT Work, a new tool designed for persistent, complex tasks, capable of handling projects for hours and automating entire workflows, from customer research to marketing asset generation. It includes "Scheduled Tasks" for repetitive, event-driven automation, even when users are offline. The tool integrates with workplace platforms like Slack, Microsoft Teams, Google Drive, and SharePoint via plugins, and its desktop version accesses local files and online resources. The coding-focused Codex app is merging into ChatGPT Work, rebranding the existing ChatGPT desktop app as "ChatGPT Classic." OpenAI warns that ChatGPT Work's complex tasks may incur higher usage costs on subscription plans, which can reach \$100 monthly with credit limits. This release coincides with the GPT-5.6 model debut, offering "stronger performance per dollar," priced at \$5 per million input tokens and \$30 per million output tokens for its highest tier.

Key takeaway

For AI Product Managers evaluating new automation solutions, ChatGPT Work offers robust, persistent workflow automation across integrated platforms. You should assess its utility by assigning a well-understood task, like budget analysis, to understand its capabilities and potential cost implications. Be mindful of the usage-based billing structure, especially for complex, long-running tasks, and utilize enterprise controls to manage spend effectively across your teams.

Key insights

ChatGPT Work enables persistent, automated workflows across applications, requiring user approval for critical actions.

Principles

Method

Automate workflows by defining a goal, allowing the agent to connect to integrated tools, and approving key actions. Schedule repetitive tasks via monitored events.

In practice

Topics

Best for: CTO, VP of Engineering/Data, Executive, Automation Engineer, Director of AI/ML, AI Product Manager

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI - Ars Technica.