Philosophers Working in or with AI Firms & Organizations (updated)
(Originally published on March 17th, 2026.)
To what extent has the development of AI over the past several years led to non-academic work for academically-trained philosophers?
[Refik Anadol, “Unsupervised – Machine Hallucinations” (detail)]
What roles are philosophers playing in the firms creating AI technologies, in the consultancies advising them, in non-profit organizations trying to keep an eye on them, in government regulation of them?
Amanda Askell of Anthropic may be the most visible philosopher in a non-academic AI role at the moment, but she is not the only one. Also at Anthropic are Joe Carlsmith, Ben Levinstein, and Jackson Kernion. Google DeepMind has Iason Gabriel, Adam Bales, Atoosa Kasirzadeh, Arianna Manzini, Julia Haas, and probably others.
The Center for AI Safety had some philosophers working for it, but it’s not clear that’s still the case. One of those philosophers, Robert Long, is now the executive director of Eleos AI, “a nonprofit organization dedicated to understanding and addressing the potential wellbeing and moral patienthood of AI systems.” The Future of Life Foundation sponsors fellowships on AI issues and at least one person whose held one of them, Paul de Font-Reaulx, is a graduate student in philosophy. Beba Cibralic is a philosophy PhD who works on AI governance and safety at the RAND Corporation.
(Thanks to Seth Lazar (ANU) for help gathering information for this post.)
UPDATE (7/6/26): Harvey Lederman is joining Anthropic AI “to work on alignment and character,” he announced recently on social media. He has been on leave from the University of Texas and has been teaching at NYU, and says he will continue to do so.
Daniel Kokotajlo is a philosopher who worked for Open AI for a while and now does other AI safety stuff I think.
Daniel was featured in the NYT last year:
https://www.nytimes.com/2025/04/03/technology/ai-futures-project-ai-2027.html?smid=nytcore-ios-share
One hopes (well, I do, for one) that AI companies are offering employment to philosophers to research basic questions that naturally arise when considering/building machine intelligence, as well as to ensure that empathy and a sense of ethics that *may* arise/evolve is not reprobate (in the sense of “missing a moral compass”).
My skeptical self worries that employing people in that discipline (philosophy) may just be a way for these companies to ease the acceptance of their “product” into the public domain.
(Side note: I worked for one of the larger Blue Cross/Blue Shield licensees in America, and once inquired as to whether there was a dedicated professional trained in ethics on staff. Nope.)
Isn’t Laurie Paul involved in some self-driving car project?
I do some light consultancy work for AI firms. Largely, this involves attempts to train the programs to reason more rigorously. I present a philosophical dilemma (usually a moral one), ask the program what we ought to do, and why. I typically have a brief back-and-forth with the program where I press it on various weak points in its argument, and write up a statement for the firm describing what I think went well, and what went poorly, with that reasoning. The actual programming of the AI is completely beyond me – my role is solely to evaluate arguments it produces.
I do the same. I wonder if it’s for the same company!
Cycorp has a long record of hiring philosophers to be “ontological engineers.” (I worked there for a year after getting my Ph.D., and I’m deeply sorry I never got business cards saying “Tim O’Keefe, Ontological Engineer.”) I don’t know how active they are now–their founder, Doug Lenat, died a few years ago–but the company has spent a long time trying to build a Good Old Fashioned AI system to model human common-sense reasoning in a giant knowledge base (Cyc), using predicate logic. (I work in ancient philosophy, not philosophy or mind or cognitive science, but I had experience teaching logic, and what they really cared about was the ability to represent statements and chains of reasoning in predicate logic.)
Yes–I was thinking of that. A long time ago some graduate students from U of Minnesota worked for Cycorp as Ontological Engineers (& yes, it’s the best job title–I always thought it was God’s). The underlying project (as I heard about it and recall…) was fascinating: it’s apparently relatively straightforward to teach expert knowledge but extremely hard to get a handle on how to teach common sense–the things we’re not explicitly taught and that don’t appear in text books.
Predicate logic?
AKA First-order logic or predicate calculus. It’s a logic of relationships that hold among individual objects in a domain, with the ability to quantify over such objects universally (“all”) and existentially (“some”). E.g., in the “Peanuts” domain of Charles Schultz, Charlie Brown and Snoopy would be objects and one could represent the statement “Snoopy is Charlie Brown’s pet” by applying the relation ‘hasPets’ to them: hasPets(CharlieBrown, Snoopy). From this you can conclude that Charlie Brown has a pet: thereExists(x) hasPets(CharlieBrown, x), or that someone has a pet thereExists(x,y) hasPets(y,x). Though Doug liked to talk about the Cyc language as though it were predicate logic, CycL was actually arbitrarily higher order, so could express things that you can’t in predicate logic, such as “there is some relationship between Charlie Brown and Snoopy” — i.e., thereExists(r) r(CharlieBrown, Snoopy) v r(Snoopy, CharlieBrown). (I worked for Cycorp for 20 years).
Ben Boudreaux, PhD from Berkeley, is also @ RAND working on AI and cybersecurity: https://www.rand.org/about/people/b/boudreaux_benjamin.html
My PhD student Ken Archer (https://www.kenarcher.org/) leads responsible AI at Microsoft. He has a BA from Tufts and a MA from CUW, his dissertation is on phenomenology and AI.
three more with NYU connections (along with five or so already mentioned above): heather whitney (ABD in philosophy at NYU), who is research counsel at anthropic and formerly copyright counsel at openAI; lisa miracchi titus (phd rutgers, postdoc NYU), who works on AI policy at meta; jorge ferreira (phd NYU), who works on chip design modeling at ASML.
It’s also worth mentioning that Lisa Miracchi Titus was a tenured associate professor at the University of Pennsylvania.
Dylan White did his PhD with me in AI Ethics at the University of Guelph and is now AI Strategy & Governance Lead with the Office of the Chief Information Officer, Government of Newfoundland and Labrador. We’ve also developed a new Master’s program in the ethics of AI with the explict aim of placing graduates in these kinds of roles
three more: patrick butlin (phd KCL), working on AI consciousness and agency at eleos AI; katja grace (former philosophy phd student CMU), who founded AI impacts; geoff keeling (phd philosophy from bristol), working on AI welfare at google.
Johnny Soraker has been with Google for a while, and Pak-Hang Wong was in industry for a bit and is now back to an academic gig. Society for Philosophy and Technology (SPT) and the Forum on Philosophy of Engineering and Technology (fPET) both have some folks who crossover (not exclusively around AI).
I am about to complete my PhD in AI ethics and governance. I am interested in roles within AI firms, but most positions I encounter require hands-on programming expertise in AI, which falls outside my training.
Michael Brent, who directs Boston Consulting Group’s Responsible AI program; Leif Hancox-Li, who works at Vijil on topics related to AI Safety / Ethics
I was a Visiting Faculty Researcher at Google from Oct 2023 – Dec 2024 working on Responsible AI and Human Centered Technology (concurrent with my jobs at Yale and then WashU), and continue to do some AI consulting for a startup.
Lisa Miracchi Titus is an AI policy manager at Meta
Two distinctions are key for mapping the stakeholder landscape: first, in terms of sectors (tech industry, international organizations, governments, etc.); second, in terms of employment status (contract-based consultancy, permanent positions, etc.).
There are a number of philosophers working AI engineering, if not building models directly. Most notable might be Alex Karp CEO of Palantir with a PhD from Goethe University. There are others working in ontology and semantic technologies like me.
William MacAskill at Forethought?
I hope AI will measure real world results against philosophical, moral, economic and political systems. Many ideas sound great until they are tested in real world environments.