Anthropic, Microsoft, Google or Meta increase exchanges with religious leaders. Researchers fear a flattened universal morality, built on Western traditions.
Fortune looks back at a discreet phenomenon: AI companies are looking to religious traditions for benchmarks to decide how their models should behave. Anthropic is the most transparent on the subject; it said in May that it had met with academics, clergy, philosophers and ethicists from more than fifteen religious and cultural groups. Philosopher Meghan Sullivan participated in one of these meetings, with Christians of several faiths, where they discussed everything from the human soul to primary school students' access to chatbots. Anthropic's Chris Olah attended the presentation of Pope Leo XIV's encyclical on AI earlier this year. Representatives from Meta, Google and Amazon have also visited the Vatican, and Microsoft is conducting its own outreach to religions.
The risk of an "average" religion
Several specialists warn against this quest for common moral principles. Anthropologist Webb Keane (University of Michigan) believes that there is a risk of privileging certain traditions and distorting others, especially those that are not Abrahamic. Paolo Carozza, a member of Meta's supervisory board, fears an artificial mixture that homogenizes beliefs in favor of Western religions. Religious scholar Tomoko Masuzawa reminds us that the very category of "religion," in the modern sense, is a nineteenth-century Western construction, foreign to many of the traditions that are claimed to be consulted.
Read the full article on the original source
Go to the source →