Source: APPSO

Just now, former Claude Code product lead Cat Wu announced she has joined OpenAI to oversee product management for core products, working alongside Tibo who manages the reset, and OpenAI President Greg Brockman also tweeted his welcome.

Interestingly, Cat Wu had just announced her departure from Anthropic on X the day before, indicating a completely seamless transition (it is estimated that A's equity has already been locked).

This marks the third departure of key members from Claude Code within two weeks.

Just ten days ago, two engineers involved in the development of Claude Code also announced their departures. One is Thariq Shihipar, who worked on Agent features, Skills, and context engineering; the other is Anthony Morris, who developed the desktop version of Claude Code.

As Anthropic seems poised for an IPO, important members of Claude Code are choosing to leave one after another, and Cat Wu's sudden shift to OpenAI adds more dramatic narratives to these two soon-to-be-listed AI giants.

Left Anthropic After 45-Fold Revenue Growth

In her resignation statement, Cat Wu shared a story.

About two years ago, Claude Code was a niche experimental tool within Anthropic. At that time, Cat Wu wanted to see what it could actually do, so she enabled the creation of a reinforcement learning (RL) environment.

She spent dozens of hours repeatedly describing her requirements to the AI and adjusting directions, yet she never manually wrote a single line of code from start to finish.

As a result, Claude Code successfully completed the task, and this experience deeply impressed her, leading her to ultimately join the Claude Code team.

Now, that initially unremarkable internal tool has become one of the world's most highly regarded AI programming products.

Anthropic revealed in February this year that Claude Code's annualized revenue run rate had already reached $2.5 billion, more than doubling since the beginning of the year.

According to the S-1/A prospectus submitted on November 2, Anthropic's quarterly revenue increased from $376 million in Q1 2025 to $16.956 billion in Q3 2026, which is about 45 times the previous figure.

Cat Wu did not mention these figures in her farewell letter, instead dedicating more space to the product itself.

Users have already started allowing AI to autonomously execute large-scale code migrations, even developing complete products from scratch. The relationship between developers and AI is shifting from "help me write a few lines of code" to "help me deliver the whole project."

Cat Wu is one of the crucial product leads in this transformation. However, she is not the traditional product manager who only handles requirement documents and develops product plans.

She transitioned to the venture capital field after working in engineering roles at Scale AI and Dagster, serving at Index Ventures. After joining Anthropic in August 2024, she initially worked as a product manager for research direction before gradually taking over Claude Code and Cowork.

Her identities as an engineer, investor, and product manager have formed a unique way of thinking.

In March this year, she stated in an official article from Anthropic that AI-era product managers should not only write PRDs but also personally use AI to build prototypes, design evaluations, and even directly participate in development. Starting in 2024, every time Claude releases a new model, she makes sure to add table functionalities to the online drawing software Excalidraw.

In Lenny's Podcast, Cat Wu mentioned that the development cycle for some products has been shortened from the previous 6 to 7 months to about 1 month, with some taking only 1 day.

Her collaboration style with Claude Code creator Boris Cherny is also representative.

Boris focuses on future technological development paths and what Claude Code can achieve next, while Cat Wu translates these ideas into tangible products and mobilizes the engineering, marketing, sales, and finance departments to complete tasks.

Her work has extended from Claude Code to Cowork, attempting to expand the capabilities of the AI Agent into document processing, data organization, and broader knowledge work.

However, the entire farewell letter did not reveal her reasons for leaving, nor did it mention her next destination. When asked if she planned to take a six-month break, she humorously replied, "post agi!"

The implication is to wait until AGI is achieved to discuss further.

In Less than 2 Weeks, 3 Key Members Departed

If Cat Wu were leaving alone, it might just be an ordinary career move in Silicon Valley. But pulling the timeline back ten days makes the story more interesting.

On October 24, Claude Code engineer Thariq Shihipar announced on X that he had left Anthropic.

He mentioned that he had worked on the Claude Code team for over a year and had witnessed the product's scale far exceeding initial expectations.

In his farewell letter, he praised Anthropic's goals and his colleagues, stating he had worked with a very talented and dedicated group and planned to take a break before starting new endeavors.

Thariq was an active engineer in the developer community for Claude Code, having contributed to features such as Skills, context engineering, and Agent workflows.

Interestingly, at the end of his farewell post, he recommended some accounts for users to follow regarding Claude Code products.

One of them included Cat Wu. But just as he was suggesting readers continue to follow Cat Wu, she also left Anthropic.

Cat Wu's departure also has a rather dramatic backstory. This is not her first time leaving Anthropic. In July 2025, according to reports from The Information, Cat Wu had previously jumped to Anysphere, which backs Cursor, alongside Boris Cherny.

Just about two weeks later, they both returned to Anthropic.

On October 23, another key member of Claude Code, Anthony Morris, also announced his departure through the X platform.

He had worked on developing the desktop version of Claude Code and believed that everyone could create the products they wanted using artificial intelligence, so he had been striving to encourage more people to use Claude Code.

At this point, the product lead for Claude Code, the creator of the desktop version, and an active engineer left the team within two weeks.

Anthony Morris represents the engineering capabilities for the desktop product, Thariq Shihipar contributed to the Agent features and developer interactions, while Cat Wu focused on higher-level product direction and inter-team coordination.

Therefore, this personnel adjustment involves not just a part of the engineering modules but affects various stages throughout the construction of the entire Claude Code product.

However, there is currently no evidence to indicate that the three left due to the same issue, nor can it be assumed that internal conflicts have arisen within the team.

Moreover, all three provided positive evaluations of Anthropic or their former colleagues in their respective farewell statements.

On the Eve of Going Public, Claude Code Faces New Challenges

This series of personnel changes occurs just before Anthropic goes public.

For an AI company about to enter the public capital market, the continued growth of its flagship product and the stability of its core team naturally become focal points for investors.

Anthropic also mentioned the risk of key talent loss in its prospectus, stating that the departure of important researchers, engineers, or product managers could adversely affect the company’s research and development pace, innovation capability, and long-term development.

However, another set of data forms a stark contrast.

The prospectus shows that Anthropic's employee size grew from about 940 at the end of 2024 to about 5,300 by the end of September 2026, and for the 12 months leading up to September 2026, the company's overall "regrettable departure rate" was only about 1%. This metric typically refers to the percentage of employees that companies wish to retain but who choose to leave voluntarily.

Why would a company with rapid personnel expansion and a low overall talent departure rate lose three key members of Claude Code in such a short span of just over ten days?

Similar personnel changes are not uncommon around the time of tech companies going public.

In August 2026, while OpenAI was preparing for an IPO, multiple senior executives left within a month, prompting co-founder Greg Brockman to become more involved in company management and to push for adjustments in the business team.

As a company transitions from startup to public entity, changes in organizational structure, management models, and operational objectives will occur.

Compared to the reasons for the departures, what draws more attention is whether Claude Code can maintain its previously impressive pace of product updates.

Cat Wu mentioned in an interview with Ars Technica that Claude Code does not have a clear long-term plan. Because the capabilities of models change rapidly, today's functionalities can be achieved with complex structures, but a few months later, models themselves can accomplish them.

This approach, driven by model capabilities and user feedback for product iteration, allows Claude Code to respond quickly to developer needs, but it also raises higher demands on the team's judgment and cooperation capabilities.

As Anthropic expands its commercialization scale, Claude Code must also address issues such as computing costs, service stability, corporate access management, and data security.

With AI lowering the barriers for software development and entrepreneurship, excellent engineers and product leads have more opportunities to independently create products, and the mobility of star talents may become more frequent.

For Anthropic, this IPO is just the beginning.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到[email protected]，本平台相关工作人员将会进行核查。