Anthropic and OpenAI hiring surge

Anthropic and OpenAI are hiring across safety, policy and alignment functions, according to their careers pages as of September 8, 2026, lending weight to recent social-media claims that non-academic roles for philosophy-adjacent candidates have become easier to spot. OpenAI’s careers site listed 780 jobs, and its recruiting page says the company is looking for people from “a wide range of disciplines and backgrounds.” Anthropic’s jobs board showed openings not only in AI research and engineering but also in public policy, public benefit and safeguards, its term for trust and safety work. (openai.com) (anthropic.com) The social posts that prompted the discussion came from X accounts TAFKAA and VersatilePhD, which pointed followers toward roles at Anthropic, OpenAI and other firms. Those posts are anecdotal, but the companies’ own job boards show the underlying hiring infrastructure is real. What they do not show, on their face, is a company asking for “philosophy majors” as such. They show something more specific: jobs built around translating human values, social norms and risk concepts into product, model and governance systems. ### Where is the evidence that this is more than a social-media vibe? (openai.com) OpenAI’s careers search page showed 780 jobs on September 8, 2026. Anthropic’s jobs board listed teams including Public Policy, Public Benefit and Safeguards, alongside AI Research & Engineering and Product Management, Support, & Operations. Anthropic’s board also showed named openings including “Research Engineer / Scientist, Alignment,” “Research Engineer / Scientist, Alignment - London,” “Anthropic Fellows Program, AI Safety & Security,” and “Anthropic Fellows Program, The Anthropic Institute (Economics & Policy).” Those postings place alignment, policy and safety work alongside core technical research rather than outside it. (openai.com) ### What kinds of jobs are these companies actually advertising? OpenAI’s “Software Engineer, AI Safety” posting says the role sits in Safety Systems and is focused on “safety, robustness, and reliability” of models in deployment. (openai.com) The job description says the engineer would build anti-abuse and content-moderation infrastructure and work with researchers to improve models’ “alignment to human values.” OpenAI is also advertising roles that look less like conventional software hiring and more like applied governance. Its “Model Policy, Chemical & Biological Risk” posting says the team “aligns model behavior with desired human values and norms” and asks candidates to build policy frameworks and taxonomies for high-risk scientific use cases. (anthropic.com) The posting says key focus areas include “catastrophic risk, mental health, teen safety and multimodal safety.” ### Why do these roles look philosophy-adjacent even when they are not labeled that way? (openai.com) OpenAI’s own language describes work that depends on defining boundaries, norms and acceptable behavior. Its model-policy posting asks where the line sits between legitimate inquiry and dangerous misuse, and says the role will turn domain expertise into “principled, operational policies” for frontier systems. Anthropic uses similar language on its careers and research pages. The company says it is building “reliable, interpretable, and steerable” AI systems, and its alignment research page is organized around work on alignment failures, oversight and model behavior. (openai.com) Those are technical programs, but they also require people who can formalize concepts, build taxonomies and reason about trade-offs in human terms. That inference is based on the job descriptions and research framing, not on a company statement that philosophy graduates are a target pipeline. ### Is this broader than Anthropic and OpenAI? (openai.com) VersatilePhD’s post also named Flagship Pioneering and Oura as recruiters of neuroscience Ph.D.s. Flagship’s roles page says it is seeking scientists, company builders and other staff across its network, while Oura’s careers page says the company wants to change how people think about health and features scientific roles including physiology-focused work. Oura’s broader business pages also show why adjacent fields matter. The company markets products and services across healthcare, employers, government, athlete performance and research, which creates demand for people who can connect scientific evidence, product claims and user-facing health concepts. (anthropic.com) ### So what can actually be said, and what cannot? The verifiable part is straightforward. As of September 8, 2026, OpenAI and Anthropic both had substantial live hiring across safety, policy and alignment-related functions, and OpenAI’s site showed 780 jobs overall. (flagshippioneering.com) The narrower claim — that philosophy majors are being directly recruited in unusual numbers — is harder to prove from company materials alone. The postings support a more careful conclusion: frontier AI companies are staffing roles that combine technical systems with questions about values, norms, misuse, oversight and governance, and those are areas where philosophy-trained or other concept-heavy candidates may have an argument for fit. (organizations.ouraring.com) Applicants looking for the next wave of openings can track Anthropic’s jobs board, OpenAI’s careers search, Oura’s careers page and Flagship Pioneering’s roles page for changes in team names, fellowship programs and safety-policy listings. (openai.com) (anthropic.com)