Tech companies turn to religion to teach AI right from wrong

Technology companies are seeking input from faith leaders and philosophers to help teach artificial intelligence how to distinguish right from wrong.

Chris Olah, co-founder of Anthropic, stated that questions regarding morality cannot be answered in a laboratory but are instead addressed by religious traditions. Olah made these comments during a Vatican presentation of Pope Leo XIV’s encyclical on artificial intelligence.

As executives debate how to develop morally good models and address the destructive potential of AI, some observers note a lack of public scrutiny regarding how these companies define religion or what they intend to gain from religious engagement.