AI companies already recruit philosophers to teach machines to “think”, but there are suspicions of “ethical washing”.
For years, Philosophy graduates were the target of jokes about the alleged lack of professional opportunities. Now, some are finding a place (and high salaries) in the biggest artificial intelligence companies, which are looking for specialists capable of helping to define how systems should respond, decide and behave.
The presence of philosophers in technology companies is no longer marginal. Amanda Askell works at Anthropic, Iason Gabriel and Henry Shevlin are linked to Google DeepMind, and Sam Altman, from OpenAI, even stated that the company used “hundreds of moral philosophers” to design rules applied to ChatGPT, recalls .
Among the problems that concern programmers are ancient dilemmas, now applied to automated systems. One of them, according to , is the call “Socratic ignorance” — the idea that wisdom begins with recognizing one’s own limits.
In the context of AI, this principle is used to try to prevent models from feigning certainty or overly agreeing with users, a behavior known as “sycophancy”.
Another central debate opposes deontological and consequentialist approaches. The first is based on strict rules — for example, not lying, not coercing and not treating people just as means to achieve ends. The second evaluates decisions based on consequences, weighing costs and benefits.
For AI creators, choosing between these models can greatly influence how a system responds to sensitive or morally ambiguous requests.
Philosophy has also entered the field of security. One example is the so-called “AI constitutionalism”in which legal or moral texts serve as a basis to guide the behavior of models. Anthropic revealed that its Claude model’s “constitution” includes references as diverse as Immanuel Kant, Apple’s terms of service, and the Universal Declaration of Human Rights — a document that has been informally dubbed the company’s “soul doc.”
Some analysts suggest that automation created by programmers could, in part, replace computer professionals, while areas such as philosophy apparently gain renewed relevance. But the rapprochement between philosophy and the technology industry is far from consensual.
Several academics fear that the rapid pace of AI will enter into conflict with the slow and critical reflection that characterizes philosophy. There is also a risk that misaligned incentives will lead to rushed or low-quality research.
The biggest criticism is the possibility of “ethics-washing”: the hiring of philosophers as a way of publicly demonstrating concern for safety and ethics, without substantially altering company practices. For critics, the presence of moral experts can function as an image strategy, especially in a sector accused of creating technologies capable of deepening inequalities, threatening democracy or being used for dangerous purposes.
