OpenAI hosted a closed-door “OpenAI Codex CTO Forum” on Monday for chief technology officers and senior engineering leaders from major Asian companies.
The event, held in Asia, brought together around 120 CTO-level leaders from large corporations and startups.
The forum focused on how AI-powered software development is shifting beyond traditional code auto-completion toward a model where AI acts as a “coding teammate” participating across the entire development process.
Codex, OpenAI’s coding agent, can generate, edit, debug and test code using natural language commands. It is also capable of handling complex feature development and code refactoring. With recent advancements in agent capabilities, Codex has evolved beyond assisting developers to directly executing specific development tasks.
Albert Yip, head of technical solutions for Asia-Pacific at OpenAI, said during the forum that AI is no longer just a productivity tool for developers but is evolving into a “delegation-based” model that can take on actual development responsibilities.
Ryu Gi-eok, a technical onboarding engineer for Codex at OpenAI, presented real-world use cases of how internal teams leverage AI agents, detailing practical development workflows within the organization.
Codex usage has been rapidly expanding globally, including in Asia, with strong growth among developers and technical users.
The launch of the Codex app for MacBook on Feb. 2 further accelerated adoption, followed by the release of a Windows version on March 5.
The Codex app recorded 1 million downloads in its first week, while weekly active users have more than tripled this year.
Kim Kyung-hoon, head of OpenAI Asia, said the forum demonstrated that AI is already moving beyond a simple development tool to becoming an active collaborator within teams.
“The event confirmed the strong interest and expectations among Asian tech leaders for agent-based development,” he said.