Workday's CTO swapped his executive position for a technical staff role at Anthropic.

      In summary: Peter Bailis, who became Workday's chief technology officer in May 2025, departed from the company last month and has accepted a position as a member of the technical staff at Anthropic, where he will concentrate on reinforcement learning engineering. This transition involves giving up a C-suite title for a more hands-on technical role, bringing Bailis to a company currently developing the HR software that Workday offers.

      An Anthropic spokesperson confirmed the hire, which was also reported by Business Insider. Bailis’s profile has been removed from Workday’s leadership pages, and his LinkedIn reflects this change. Before joining Workday, Bailis was the vice president of engineering at Google Cloud, where he headed AI initiatives for data, including Conversational Analytics, NL2SQL, and retrieval-augmented generation for structured data. Previously, he founded and led Sisu Data, a decision intelligence firm, and served as an assistant professor of computer science at Stanford University, where he co-led the DAWN project focused on data-intensive applications. His research background in data systems and recent enterprise AI experience at Google Cloud make him an atypical candidate for a reinforcement learning role, indicating Anthropic's interest in engineering that connects model training with practical product development.

      From an executive role to a contributor position

      The title Bailis has taken at Anthropic, member of technical staff (MTS), is the standard engineering designation at both Anthropic and OpenAI and is applied across research and engineering irrespective of rank. OpenAI president Greg Brockman has publicly discussed this structure, stating that the labs aim to avoid categorizing individuals strictly as researchers or engineers, treating the MTS title as indicative of a flat technical hierarchy rather than a downgrade. In reality, the title covers a broad spectrum of seniority and compensation: MTS base salaries at Anthropic range from approximately $300,000 to $405,000, while at OpenAI, the range extends from roughly $210,000 to $530,000, with equity grants potentially elevating total compensation into the seven-figure range for senior roles.

      For Bailis, the trade-off is clear. A CTO position at a company like Workday, with around $8 billion in annual revenue and 18,000 employees, carries a level of institutional authority that an individual contributor role at a cutting-edge AI lab does not. However, it offers direct engagement with the models being developed, the training decisions being made, and the research agenda being determined, without the layers of organizational management separating a CTO from the actual work. Anthropic, which now reports a revenue run rate surpassing $30 billion and serves over 1,000 enterprise clients each spending more than $1 million annually, is now beyond being merely a research lab that generates revenue — and for a technically ambitious executive, the complexity of the engineering challenges it now faces is appealing in itself.

      Bailis's value to Anthropic now

      The Anthropic spokesperson confirmed that Bailis will focus on reinforcement learning engineering. This role encompasses a wide remit at a company where reinforcement learning from human feedback is essential for model alignment and where such training methods are being used for agent behavior and tool use. Beyond his general engineering skills, Bailis brings extensive knowledge of the enterprise software landscape that Anthropic is now entering as a product company rather than just an API provider.

      Specifically, The Information noted that Bailis’s entry coincides with Anthropic's initiative to develop HR applications. Anthropic has already introduced Claude plugins for HR purposes, such as generating job descriptions, onboarding materials, and offer letters. Bailis has spent recent years immersed in the enterprise HR and finance software domain, first at Google Cloud, where he created AI products integrated with structured enterprise data, and then as Workday’s CTO, leading the company’s strategy for agentic AI. This domain expertise is now directed towards enhancing Anthropic’s enterprise product roadmap from an insider perspective.

      The competitive irony is noteworthy. Workday’s CEO, Carl Eschenbach, publicly stated in February 2026 that Anthropic, Google, and OpenAI all utilize Workday’s software internally. The company that previously licensed technology from Workday has now recruited its CTO to develop products that will compete for the same enterprise HR resources.

      Anthropic's enterprise ambitions

      Bailis’s recruitment aligns with Anthropic's aggressive development of enterprise distribution infrastructure over the past months. In early March, Anthropic launched a marketplace for Claude-powered enterprise software, allowing firms with committed API expenditures to purchase third-party applications developed on Claude, with Anthropic foregoing the typical revenue share that cloud hyperscalers collect. Initial launch partners included Snowflake, the legal AI firm Harvey, and developer platform Replit.

      Shortly after, Anthropic pledged $100 million to establish a Claude Partner Network with Accenture, Deloitte, Cognizant, and Infosys as key partners, formalizing consulting relationships aimed at accelerating Claude deployments within the world's largest enterprises. The company has indicated intentions to