AI companies are hiring philosophers

One of the unexpected winners of the investment boom in artificial intelligence has been philosophy. As technology companies look for expertise in ethics, reasoning and AI safety, demand for philosophy graduates and academics is rising fast.

What the numbers say#

According to data from the Federal Reserve Bank of New York, unemployment among philosophy graduates in the United States stood at 5.1 per cent in 2024, against 7 per cent for computer science graduates. Ten years ago humanities students were being told to learn to code if they wanted a job; the picture appears to have reversed.

"A haemorrhage"#

Luciano Floridi, professor of philosophy at Yale University, uses the word haemorrhage for the movement of academics out of philosophy departments and into the technology sector. Companies such as Anthropic, Google DeepMind, IBM and OpenAI now either employ philosophers directly or build philosophical principles into the design of their models.

Where does philosophy sit in a model?#

The sources these companies draw on are surprisingly classical: Socratic questioning, and the value theories of Kant and Locke. The approaches are used to improve a model's reasoning, to test assumptions, to reduce the invented answers known as "hallucinations" and to guard against harmful behaviour.

At the centre of the debate is an old distinction in moral philosophy: should an AI follow fixed principles, or weigh consequences case by case? Frameworks such as constitutional AI represent the first path; approaches based on a calculus of utility represent the second.

Why now?#

As artificial intelligence settles into health, transport, law and defence, questions of justice, responsibility and accountability move inside technical decisions. That turns philosophy — long described as the subject that does not lead to a job — into a directly applied field.