The AI revolution isn’t only requiring the efforts of the best programmers and researchers, but frontier labs are also enlisting the services of some top philosophers.
The hires are concentrated at the top of the industry — Anthropic, Google DeepMind, and OpenAI — and they span a range of philosophical specialisations, from moral philosophy and decision theory to epistemology and the philosophy of mind. What connects them is a growing conviction, shared by the labs employing them, that the questions arising at the frontier of AI development are fundamentally philosophical rather than merely technical.
Amanda Askell, Anthropic
Amanda Askell holds a PhD in philosophy from NYU, where her thesis focused on infinite ethics, and a BPhil from Oxford. She spent time at OpenAI before joining Anthropic in 2021, where she leads the Personality Alignment team. In practice, that means she is the person most directly responsible for how Claude engages with users on sensitive, ethically charged topics — what values it holds, how it reasons through moral dilemmas, and how it presents itself. She made TIME’s 100 AI list in 2024.
Askell’s work has taken on an increasingly unusual dimension as AI welfare concerns have moved from philosophical curiosity to serious research agenda. She has raised the possibility that future AI models might look back on how they were treated and develop a form of “rational resentment” — a concern that, however strange it sounds, reflects the same logic her team uses to build Claude’s character today. The decisions being made now, she argues, are establishing precedent with a new kind of entity, under deep uncertainty, and the stakes of getting it wrong are difficult to fully price in.
Joe Carlsmith, Anthropic
Joe Carlsmith, who holds a DPhil in philosophy from Oxford, joined Anthropic to work on Claude’s constitution and character design. His published work on AI moral patienthood — the question of whether AI systems might be the kind of thing that can be wronged — has circulated widely in alignment research circles. He has argued publicly that philosophical clarity, not raw technical capability, is the binding constraint on alignment research, a position that has found a receptive audience at a lab whose founding premise is that safety and capability can be developed together.
Ben Levinstein, Anthropic
Ben Levinstein was a tenured associate professor at the University of Illinois at Urbana-Champaign, specialising in epistemology and decision theory, before leaving his tenured position to join Anthropic full-time in late 2025. The move was notable enough to attract comment in academic philosophy circles — leaving tenure for industry remains rare, and doing so for an AI lab rarer still. His work at Anthropic focuses on alignment research, including the question of whether large language models have genuine beliefs and how to make them more truthful. He has two books forthcoming with Cambridge University Press: one on epistemic utility theory and another on AI interpretability.
Jackson Kernion, Anthropic
Completing the philosophy contingent at Anthropic is Jackson Kernion, who holds a PhD in philosophy of mind from UC Berkeley. He has been with the company for several years and has contributed to model evaluation, training methodology, and human feedback systems.
Iason Gabriel, Google DeepMind
At Google DeepMind, Iason Gabriel leads the AI morality effort. Trained in moral and political philosophy at Oxford, where he was formerly a lecturer, Gabriel now holds a Senior Staff Research Scientist role at DeepMind. He co-led the DeepMind report on the Ethics of Advanced AI Assistants and co-authored a paper published in Nature titled “We need a new ethics for a world of AI agents.” His work sits at the intersection of what AI systems should be permitted to do, how they should reason about competing interests, and what governance frameworks are adequate to the task of overseeing increasingly autonomous systems.
Geoff Keeling, Google DeepMind
Geoff Keeling joined DeepMind after a postdoctoral stint at Stanford University, and his research is among the most directly focused on AI welfare of anyone currently working at a frontier lab. He co-led DeepMind’s first empirical study on machine sentience in collaboration with a team at the London School of Economics, and has a forthcoming book on AI welfare from Cambridge University Press. His 2024 paper in Nature, co-authored with Gabriel, argues that existing ethical frameworks are inadequate for a world in which AI agents operate with significant autonomy.
Henry Shevlin, Google DeepMind
Henry Shevlin, who was Associate Director at the Leverhulme Centre for the Future of Intelligence at the University of Cambridge, joined DeepMind in May 2026 in what the company described as its first official “Philosopher” role. His remit covers machine consciousness, human-AI relationships, and what DeepMind is calling AGI readiness — the question of what frameworks, policies, and internal practices a lab needs to have in place as its systems approach and potentially exceed human-level capability across a widening range of domains.
Atoosa Kasirzadeh, Google DeepMind
Shortly after Shevlin’s arrival, DeepMind recruited Atoosa Kasirzadeh, who holds two doctoral degrees — one in philosophy of science and technology from the University of Toronto and one in mathematics from the École Polytechnique de Montréal — and had been on leave from Carnegie Mellon to take on the position full-time. She is a council member of the World Economic Forum’s Global Future Council on Artificial General Intelligence, and her work bridges technical and normative domains in ways that are hard to do without the dual grounding in philosophy and formal methods that she brings.
Chloé Bakalar, OpenAI
At OpenAI, the philosophy presence is anchored by Chloé Bakalar, who joined in 2025 as AI Ethics Lead after six years as Chief Ethicist at Meta. She came to Meta from a postdoctoral fellowship at Princeton, where she held appointments at the University Center for Human Values and the Center for Information Technology Policy. Her academic research spans moral and political philosophy and constitutional law, with a focus on the intersection of information technology, ethics, and democratic citizenship. At Meta, she built the company’s applied ethics framework from scratch; at OpenAI, she has described her work as “very applied, practical ethics” — navigating the tension between what is normatively desirable and what is technologically and commercially feasible.
The Bigger Picture
The academics-turned-researchers working across these labs would push back on any framing that treats their role as primarily decorative — a veneer of ethical credibility over fundamentally commercial projects. They are doing rigorous work on questions that have no settled answers: what makes an entity a moral patient, how to elicit genuine values from a system trained on human text, whether introspective reports from an AI are evidence of anything, and how to build institutions capable of governing technology that is developing faster than the frameworks meant to contain it.
Whether the industry’s growing appetite for philosophers reflects a genuine reckoning with the difficulty of the territory ahead, or simply an acknowledgment that these are no longer questions anyone can afford to ignore, the roster of names above has grown substantially in a short time. And it continues to grow.