
OpenAI released ChatGPT Work to bring AI agents to non-engineers.
It handles tasks across email, Slack, and other apps.
The tool costs $20 monthly and targets white-collar workflows.
What happened
OpenAI released ChatGPT Work last month, a version of its Codex coding tool for non-engineers. It connects AI agents to email, Slack, Notion, and more, and is available on the lowest subscription tier at $20 a month.
Why it matters
The product aims to bring AI agents to white-collar workers beyond software engineers, a market OpenAI sees as crucial for growth. An OpenAI-backed study found only 17% of organizational and less than 1% of individual subscribers used Codex, versus 98% of employees, highlighting the adoption gap.
What to watch
OpenAI faces challenges in making the tool intuitive and reliable for non-technical users, and in evaluating work that isn't as measurable as code. The company uses its GDPval benchmark, drawn from 44 occupations, to guide development.
Ask the AI about this article →
ChatGPT Work is OpenAI's attempt to expand AI agents beyond software development, a field where tools like Codex have seen strong adoption. The company's internal study shows a stark contrast between employee usage (98%) and subscriber usage (17% for organizations, under 1% for individuals), underscoring the challenge of making AI agents appealing to a broader workforce. OpenAI's engineers are focused on improving the 'harness'—the software that connects the model to tools and instructions—to make the experience more intuitive, much like how graphical interfaces replaced command lines for everyday users.
The product's commercial logic is clear: agents that work longer consume more tokens, increasing revenue per user. To succeed, OpenAI must not only refine the technology but also convince cautious users to grant access to sensitive data, a barrier the article's author personally grappled with. The company also faces competition from vertical-specific rivals like Harvey and Clay, which tailor AI to law and sales, respectively. OpenAI's path forward relies on learning from early adopters and continuously improving the product based on feedback, as the broader goal is to move AI from answering questions to autonomously completing complex tasks for everyone.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Israeli startup DataAgent Ltd
SK Hynix presented a custom HBM concept at SEMICON Taiwan 2026, where compute functions are placed in the base…

Taoyuan is positioning itself as a northern hub for AI data centers (AIDC), citing the Tatan area and an LNG c…

The U.S. Department of Defense announced on August 31 that it has deployed ChatGPT Mil, a customized version o…

Nvidia reported earnings that were both remarkable and boring, reflecting its focus on avoiding a consolidated…

Anthropic has agreed to a $35bn cloud-computing contract with Lambda, a Nvidia-backed cloud provider
