Why Have AI Companies Started Hiring Philosophers?

What do they do?

“Philosopher” is a strange title to find on a corporate org chart. So when Google DeepMind hired Henry Shevlin under that title, it drew notice. Inside frontier AI labs, philosophers had been working on questions of alignment, governance, and long-term risk for years. Their presence was old news. What was new was the title on the door. That shift is a useful signal of where the hardest problems in AI now sit.

Anthropic, OpenAI, and DeepMind have each built internal capacity for normative reasoning, often quietly. At Anthropic, Amanda Askell leads the effort to shape Claude, the company’s language model, through a written constitution: a document articulating the principles by which the model should respond when requests pull in different directions. OpenAI is less explicit in its philosophical titles but is engaged with closely related questions; its public work on alignment and model behavior involves frameworks for how systems should respond in situations that no one can fully anticipate. Iason Gabriel, a political philosopher at DeepMind whom I once heard lecture as a student at Oxford, has spent years on the value alignment of AI systems.