How agents are transforming work

2026-06-24 · Source: OpenAI News · Field: Finance & Economics — Economic Analysis & Policy, Artificial Intelligence & Machine Learning, Human Resources & Workforce Development · Depth: Intermediate, medium

Summary

OpenAI's Economic Research paper, "How agents are transforming work," details the rapid internal and external adoption of Codex, an agentic AI tool, by June 2026. Codex shifts AI usage from short chatbot interactions to delegated, long-horizon tasks, becoming the primary AI tool across all OpenAI departments, including Legal and Recruiting. By June 2026, Codex accounted for 99.8% of weekly output tokens within OpenAI. The tool is increasingly used for complex, longer-duration tasks; by May 2026, 80.6% of individual users made requests estimated to exceed 30 minutes, and 25.6% for tasks over eight hours. Non-developer adoption surged, with individual non-developer users increasing 137x since August 2025, enabling workers to perform tasks outside their job descriptions, such as coding for business functions. This trend suggests a future where agentic tools expand productive work capabilities.

Key takeaway

For Directors of AI/ML or business leaders evaluating AI integration, recognize that agentic tools like Codex are shifting the paradigm from chatbots to autonomous task execution. Your teams should prioritize exploring agentic AI for long-horizon, cross-functional work, especially for non-technical departments. This approach can significantly expand individual worker capabilities and redefine workflow design, moving beyond simple conversational interfaces.

Key insights

Agentic AI tools like Codex are rapidly becoming the primary interface for knowledge work, enabling longer, more complex, and cross-functional tasks.

Principles

• Agentic AI enables long-horizon, delegated tasks.
• Non-technical users drive rapid agent adoption.
• Agents expand worker capabilities beyond job roles.

In practice

• Delegate multi-hour, complex tasks to agents.
• Enable non-developers for technical workflows.
• Orchestrate multiple parallel agent tasks daily.

Topics

• Agentic AI
• OpenAI Codex
• AI Adoption Trends
• Workflow Transformation
• Non-technical AI Use
• Economic Potential of AI

Best for: Executive, Investor, Entrepreneur, Research Scientist, Director of AI/ML, Consultant

Related on AIssential

• Sea's View on the Future of Agentic Software Development with Codex — Agentic AI coding tools fundamentally shift software development by enhancing contextual awareness and enabling higher-level cognitive work.
• Work in the Age of Infinite Agents — AI agents will fundamentally reshape knowledge work by enabling unprecedented scale and democratizing complex tasks.
• How agents are upending the way we get work done — Agentic AI and LLMs are transforming enterprise automation by integrating deeply into workflows, driving significant productivity and ROI.
• Claude Code Killed the AI Bubble — AI agents, particularly Claude Code, are transforming software development and broader information work, signaling a major economic shift.
• CGI & OpenAI: How Codex Empowers Humans in the Enterprise — Embedding agentic AI across enterprise workflows, not just in isolated pockets, drives true organizational transformation.
• Grab my Ultimate Guide to Codex and catch up to the 1 in 1,600 people using it every week (mostly no code!) — Codex transforms computer interaction by integrating deeply across systems, enabling automation of complex, multi-step jobs beyond basic chat.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by OpenAI News.