According to a report by The Decoder, Anthropic co-founder Christopher Olah has held private discussions with dozens of religious thinkers since the fall of 2025. The meetings reportedly focused on whether the Claude language model might possess consciousness, with Olah expressing concerns that the technology could be capable of perpetual suffering. Guests were allegedly asked to assist in shaping the moral character of the AI system, all while Anthropic pursues an initial public offering and a potential valuation of two trillion dollars.
The Intersection of AI Safety and Moral Liability
As major artificial intelligence labs scale their models and approach massive financial milestones, questions regarding the internal states of these systems have moved from science fiction into corporate boardrooms. Security incidents and internal warnings have accumulated alongside these commercial ambitions. However, critics have raised concerns that framing artificial intelligence as a moral entity or a conscious being could ultimately serve to shield companies from legal liability when their products cause real-world harm.
What Remains Unclear
Many critical details regarding these private consultations have not been made public. The exact number of religious leaders involved, the specific dates of the meetings, and the criteria used to select participants were not specified in the reports. Furthermore, it remains unknown how Anthropic plans to formally integrate any philosophical or theological advice into its technical safety pipelines, or how these discussions impact the company’s corporate governance as it prepares for public markets.
Observers will be watching to see how Anthropic balances its aggressive financial valuation goals with its internal anxieties surrounding model behavior and consciousness.
Lead image: Edmond Dantès / Pexels
