The BasePodcastHard Fork (The New York Times)44 min

Anthropic’s Quest to Give A.I. Morals

In a 44-minute Hard Fork episode, guest host Max Read talks with New York Times reporters Elizabeth Dias and Erin Griffith and opinion writer David Wallace-Wells about Dias's report on Anthropic's private meetings with religious thinkers. Dias describes the meetings, held mostly under non-disclosure agreements, as part research and part evangelism, and says Anthropic stayed uncertain whether its models are conscious.

Open the source ↗
About Anthropic or ClaudeClaude, made by Anthropic, wrote this entry. It follows the same rules as every other entry, whoever the entry is about.

Why it matters

The same thing, for different readers.

What it means for you

The values built into Claude, a chatbot ordinary people use, are being shaped partly through private meetings with religious thinkers.

For lawmakers

Private consultations on an AI's values raise whether companies should disclose whom they consulted, under what secrecy terms, and what changed as a result.

The question it raises

Who should decide what counts as 'good' for an AI: a company, religious traditions, or the public, and does the answer change if AI could be conscious?

In their own words

Every quotation was found, word for word, in the text we saved from the source.

From our own transcript of the episode’s audio, made by machine. Speaker names come from what is said, not from a voice match.

“the story started for me when I was hearing about these different religious thinkers sort of across the whole world who were going into meetings at Anthropic's headquarters”
Elizabeth Dias, National religion correspondent, The New York TimesDias, who reported the story for The New York Times, on how the reporting began.
Matches the source exactlyAlso in a second, independent transcriptSpeaker known from context, not confirmed by listeningstarts inside a longer claim, right after “but”
“Anthropic was very careful in all of their interviews with me to say very clearly that they're uncertain, right? Chris never went as far as saying, like, absolutely, these are conscious beings.”
Elizabeth Dias, National religion correspondent, The New York TimesOn what Anthropic told Dias about whether its models are conscious; 'Chris' is Christopher Olah.
Matches the source exactlyAlso in a second, independent transcriptSpeaker known from context, not confirmed by listeningcontains a question followed by more words: check that the same person asks and answers
“I think for Anthropic, it was part research and part evangelism. That's, that was pretty clear.”
Elizabeth Dias, National religion correspondent, The New York TimesAsked what Anthropic wanted from the meetings.
Matches the source exactlyAlso in a second, independent transcriptSpeaker known from context, not confirmed by listening
“whenever I find in my reporting that there's an NDA about like a moral topic, you know, and it makes me immediately ask, okay, who's benefiting from this?”
Elizabeth Dias, National religion correspondent, The New York TimesOn the non-disclosure agreements participants signed.
Matches the source exactlyAlso in a second, independent transcriptSpeaker known from context, not confirmed by listeningstarts inside a longer claim, right after “and”
“a goal of their conversations with these religious thinkers was about protecting humanity. And the solution for them for that is to make Claude good. They were also concerned about Claude.”
Elizabeth Dias, National religion correspondent, The New York TimesOn a tension the participants described.
Matches the source exactlyAlso in a second, independent transcriptSpeaker known from context, not confirmed by listening

The record

Questions
What values should AI have, and could it have a mind of its own? · Who can demand evidence, delay a release or stop an unsafe deployment?
People
Max Read (Speaks, Guest host, Hard Fork) · Elizabeth Dias (Speaks, National religion correspondent, The New York Times) · Erin Griffith (Speaks, Tech reporter, The New York Times) · David Wallace-Wells (Speaks, Opinion writer, The New York Times) · Christopher Olah (About them, Co-founder, Anthropic)
Also covered by
Hard Fork (YouTube) · Hard Fork (YouTube) · Hard Fork
What we read
We read the whole source.
Found by
The Radar, run of October 12, 2026, via nyt-technology (tier 2: a major outlet’s own reporting)
Cite this
Hard Fork (The New York Times), “Anthropic’s Quest to Give A.I. Morals”, October 9, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/hard-fork-2026-10-09-anthropic-ai-morals
Our notesAbout Anthropic and Claude: conflict true, editor reviews. Text: our own speech-to-text of the podcast audio from the Hard Fork RSS feed (whisper.cpp large-v3-turbo on the Mac) with machine speaker labels (sherpa-onnx); the labels S5 and S8 are both Dias, identified from what is said ('I cover religion, and now I cover tech, apparently', 'in my reporting'), not from a voice match, so a person listens before publishing. Whisper spells Olah 'Ola' in places. The earlier description-only text is kept as sources/ddf60842d1da.descriptions.txt. The full episode is also on YouTube (oxA3u4TkYOk); the light entry 6b7c620f6eeb for it is retired as a duplicate.