Cat Wu announced on November 5, 2026 that she is joining OpenAI to lead product management for Core Products.
Wu said she respected OpenAI for showing that scaling works, bringing AI to more than a billion people and treating great products as central to its mission. She named Thibault Sottiaux, OpenAI’s product-management team and employees across the company as people she was excited to work with.
Her stated objective is to build agents that support users and their teams, keep users in control and give them more time for the people and work they love. She also invited continued feedback.
Claude Code Background
The move follows Wu’s November 3, 2026 announcement that she had left Anthropic. She joined the Claude Code effort two years earlier, when it was a small internal tool, after asking it to build a reinforcement-learning environment and steering it for tens of hours without manually writing code. In her account, usage developed from manual approval of permission prompts and Claude use focused on answering questions and writing tests toward default auto-approval, large migrations and building new products. She credited the Claude Code, Cowork and Tag teams, along with users whose feedback shaped what they shipped.
Persistent Agents and User Control
Sottiaux is named as OpenAI’s Core Products Lead in the June 11, 2026 announcement of OpenAI’s planned Ona acquisition, which would bring Ona’s secure cloud execution and orchestration technology into Codex. OpenAI reported at the time that more than 5 million people used Codex each week, up 400% from earlier in 2026, and that the tool had expanded beyond software development into research, analysis, building and automation. The announcement connected its next phase to work extending over hours or days rather than minutes.
The planned Ona integration would provide persistent environments where agents can access tools, systems and context over time, including work continuing inside a customer’s cloud environment when laptops are closed. OpenAI’s design is intended to let users check progress, provide direction, make decisions and review results remotely rather than remaining tied to the device or session where a task began.
For organizational deployments, the documented requirements are control over where agents run, what they can access, how credentials are scoped, how activity is logged and how work moves through review. Ona’s customer-controlled execution model would let agents operate inside an organization’s own cloud environment while OpenAI provides intelligence and orchestration. Those controls give Wu’s emphasis on user control a concrete enterprise context: delegation depends on the agent’s capabilities and on the boundaries around its work.
Greg Brockman, identified as OpenAI’s president and co-founder in OpenAI’s October 21, 2026 Google Cloud announcement, welcomed Wu after her post. Sottiaux also welcomed her and said he looked forward to their work together.
Wu’s career move connects complementary approaches to the same product problem. Her Claude Code account describes a progression from approving individual actions to delegating large tasks, while OpenAI’s documented agent plans combine persistent work with remote review and organizational controls. Her appointment brings that experience to Core Products, with the product challenge being how to expand delegated work while preserving users’ ability to direct, inspect and govern it.
Credit: Source link
