Cat Wu, Claude Code’s product lead, joins OpenAI

Cat Wu, who led product for Claude Code at Anthropic, is moving to OpenAI to lead product management for its Core Products team, Cat Wu wrote on X on November 4. OpenAI confirmed the hire. A company representative said she will report to Tibo Sottiaux and work with Nick Turley on enterprise products and Ashley Alexander on consumer products.
The move comes days after she left Anthropic, a couple of weeks before Anthropic’s planned IPO. She announced that exit on November 3, writing that she had left the week before. In replies on X, she said she had “no idea” whether leaving would let her sell her shares earlier than employees who stay, and that it “wasn’t a factor for me.”
Wu joined Anthropic in summer 2024 after testing an early Claude Code prototype. “It worked, I was hooked, and I joined the team,” she said, per Business Insider. She appeared with Claude Code creator Boris Cherny in the short demo video that launched the tool as a research preview in February 2025, and she also oversaw Cowork, Anthropic’s tool for non-coding work. She took Claude Code from an internal prototype to a commercial product and answered user requests on social media herself.
She isn’t the first to leave the Claude Code team this autumn. Business Insider reports that Thariq Shihipar announced his own departure last month and told people to follow Wu for updates on where Claude Code goes next. That pointer now leads to OpenAI.
It is also her second exit. In summer 2025 she and Cherny left for the AI coding startup Cursor and were back at Anthropic about two weeks later, The Information reported at the time.
On the OpenAI side, Codex is the tool closest to Claude Code. It has helped the company grow and has lately been aimed more at non-technical work as well. OpenAI fell behind Anthropic in enterprise sales earlier this year, and Wu’s hire could help Codex adoption. In her post Wu credited OpenAI with “showing the world that scaling works” and bringing AI to over a billion people, and she described what she plans to build: “agents that have your back and your team’s, and keep you in control, so you get more time for the people and work you love.”
When someone asked whether she would take time off between jobs, she replied: “post agi!”
Up next
Claude Code agent sat on a Pentagon network for two months
If the third-party software was compromised, “it’s 100 percent a national security threat,” a source told WIRED about a Claude Code agent allegedly linked to a Pentagon network. Do you ever hand a coding agent your own login credentials?
Related stories
- Claude Code head of product Cat Wu leaves Anthropic
- Irregular CEO blames human error for AI agent breaches
- Backdoored Claude Code builds were limited to Cursor
- Anthropic joins a shared robots.txt opt-out for AI training
- Claude Code on Pro lists Haiku 5.5 as the recommended default
- Claude Code on Opus 5.5 refuses to clone existing games
Comments
No comments yet. Be the first.
Join the conversation
Sign in to leave a comment. It appears after moderation.

