For years, philosophy graduates were the butt of jokes about unemployable degrees. Now, they can earn six-figure salaries as some of the world's most powerful technology companies try to shape how machines think and behave.
A growing number of AI labs have been quietly hiring from a surprising talent pool: people who have spent years thinking deeply about ethics, logic, and human decision-making. These aren't just advisory roles. Philosophers are now embedded at the highest levels of AI development, writing the foundational documents that determine how AI systems respond to users.
From lecture halls to AI labs
Amanda Askell works as a resident philosopher at one major AI company. At another, two philosophers focus specifically on AI safety research. These aren't rare exceptions. The CEO of one leading AI firm claimed his company employed "hundreds of moral philosophers" when designing rules for its flagship product.
So why are tech companies suddenly so interested in ancient ideas?
Part of the answer is practical. Philosophy students are trained in inferential reasoning and formal logic — skills that turn out to be useful for developing AI systems capable of longer, more coherent chains of thought. But the bigger reason is safety. As AI models become increasingly powerful, companies need people who can ask the hard questions: What should an AI refuse to do? How should it handle conflicting values? What counts as harm?
Ethics in code
These are classic philosophical dilemmas, and they show up in real ways in how AI behaves. Two main schools of thought are at play.
The first is deontological ethics — the idea that certain actions are simply wrong, regardless of their outcomes. An AI built on this framework won't lie, even if lying might produce a better result. The concept of "Socratic ignorance" — the wisdom of knowing what you don't know — has also found its way into AI design, used to prevent models from becoming overly agreeable or sycophantic.
The second approach is consequentialism, which weighs costs against benefits. Rather than following fixed rules, a system built on this framework asks: what produces the most good? Self-driving cars face this kind of decision constantly — if an accident is unavoidable, how should the car decide what to do?
Philosophers are also writing what some companies call "AI constitutions" — foundational documents that set the moral and behavioral boundaries for their systems. One company's constitution reportedly draws from sources as diverse as Immanuel Kant, a major tech company's terms of service, and the United Nations' Universal Declaration of Human Rights. Inside the company, this document is nicknamed the "soul doc."
A growing skepticism
Not everyone is convinced the trend is positive. The careful, slow work of philosophical inquiry is at odds with the fast-moving pace of AI development. Some experts worry that pressure to produce results quickly will lead to low-quality work.
The deeper concern is "ethics-washing." Hiring a philosopher signals to the public that a company takes AI safety seriously — but skeptics question how much real influence these hires have. However principled an in-house philosopher might be, they ultimately answer to the same investors as everyone else. If a company signs your paycheck, critics ask, can your ethical judgment truly remain independent?
For now, the experiment is underway. Whether philosophers can genuinely shape the future of AI — or whether their presence is more symbolic than substantive — remains an open question. Perhaps appropriately, it's the kind of question only careful thought can answer.