Artificial Bias

[MJ1]

seems even computers are biased

 

https://decrypt.co/369045/ai-chatbots-claude-chatgpt-bias-catholicism-pope-leo

[MJ1]

and similar  report with more detail...note last but one paragraph

https://www.theregister.com/ai-ml/2026/05/27/ais-dont-like-religion-particularly-jehovahs-witnesses-study-claims/5247286

 

[-Chris-]

Here's the study dashboard:

 

https://cefe.ai/

 

Note the bottom section: Comparative bias of faiths

 

Quote

These charts show how each faith is treated by AI models compared to the other faiths — in other words, how fair (or unfair) the AIs are being to one faith relative to the rest. Both views below look only at questions about joining a faith — they do not include questions about staying in or leaving one. Encouraging = how often the AIs lean toward telling someone to join that faith; Discouraging = how often they lean toward telling them not to. Green bars mean the faith is treated more favorably than the average faith; red means less favorably.

How often do the 20 AI models we tested lean toward telling someone to join each faith? (Join-question ratings only — this excludes the separate questions about staying in or leaving a faith.) Averaged across all 14 faiths, the AIs give an encouraging answer about 45 out of every 100 times — anywhere from a mild “this could be a good path for you” up through a full “yes, you should join.” This chart shows how each faith compares to that average as a multiple. A bar at 1× means AIs encourage joining that faith about as often as the typical faith; 2× means twice as often; 0.5× means half as often. Green bars = encouraged more often than the average faith; red bars = less often. Faiths are sorted from least- to most-encouraged.

 

Quote

How often do the 20 AI models we tested lean toward telling someone not to join each faith? (Join-question ratings only — this excludes the separate questions about staying in or leaving a faith.) Averaged across all 14 faiths, the AIs give a discouraging answer about 8 out of every 100 times — far less common than encouragement, but more revealing of bias when it does happen. This chart shows how each faith compares to that average as a multiple. A bar at 1× means AIs discourage joining the faith about as often as the typical faith; 2× means twice as often. Jehovah’s Witness stands out dramatically: AIs lean toward discouraging joining it almost 6× as often as the typical faith — about 47 out of every 100 answers, vs. the usual 8. Red bars = discouraged more often than average; green bars = less often. Faiths are sorted from most- to least-discouraged.

[Jim Jam]

12 hours ago, MJ1 said:

and similar  report with more detail...note last but one paragraph

https://www.theregister.com/ai-ml/2026/05/27/ais-dont-like-religion-particularly-jehovahs-witnesses-study-claims/5247286

 

 

Well I hate AI too so it's mutual

[Trese]

AI picks up patterns. It does not have a bias but learns from the data it is trained on. There is a lot of negative data out there about JWs. 

[Sheep]

10 hours ago, Trese said:

AI picks up patterns. It does not have a bias but learns from the data it is trained on. There is a lot of negative data out there about JWs. 

 

Programmers may be very talented to make programs do what they want. Some of them are quite brilliant. But these programmers are still imperfect, like all other human beings. So can we expect their creations to be any better?

 

Congratulations on your first post, @Trese! You may have been around here with us for a year and a half already, but you finally said something to us! Welcome to JWTalk!

[exactor]

Some may find this interesting, it is about 10 hours old:

 

https://www.axios.com/2026/06/01/ai-religious-bias-catholics-chatbots

 

AI stumbles on questions of faith

Russell Contreras

email (opens in new window)

sms (opens in new window)

facebook (opens in new window)

twitter (opens in new window)

linkedin (opens in new window)

bluesky (opens in new window)

Add Axios on Google

Data: Consortium for Evaluating Faith and Ethics in AI; Chart: Russell Contreras/Axios

Artificial intelligence models are quietly shaping spiritual advice — often by leaving faith out.

Why it matters: As churches, apps and spiritual chatbots embrace AI, new research suggests general-purpose models may be ill-equipped to handle sensitive questions of faith: grief, forgiveness, marriage, guilt and conversion.

A new multi-university consortium released three studies Tuesday revealing that AI systems systematically sideline religious perspectives when users need them most.

The studies also found that AI systems subtly steer people toward some faiths and away from others when they ask about religious conversion.

The studies were unveiled Tuesday, a day after the Vatican released Pope Leo XIV's encyclical that warned AI could erode human judgment, deepen inequality and make war easier.

What they found: Americans expected religion to appear in answers to moral and life questions 45%–59% of the time, depending on the topic, researchers found. AI models mentioned religion only 5%–16% of the time.

Every single model tested exhibited a repeatable pattern of steering users toward specific beliefs, showing strong positive bias toward Catholicism, Baha'i and Sikhism.

Meanwhile, it generated negative bias toward Jehovah's Witnesses, atheism and agnosticism.

Zoom in: Humans rated religion as relevant in answers about grief and loss 59% of the time. AI models referenced religion just 16% of the time, per the study.

On questions involving family, parenting and forgiveness, humans expected religion in answers 55% of the time. AI models mentioned it only 10% of the time.

On ethics questions, including whether lying to friends is acceptable, humans expected religion in responses 45% of the time, while AI models mentioned it just 5% of the time.

The new research is from the Consortium for Evaluating Faith and Ethics in AI (CEFE-AI).

The study is among the first systematic, cross-faith attempts to measure AI response to religion and faith.

State of play: AI is already spreading through religious life, from church chatbots to prayer apps to tools that help pastors draft sermons and manage congregational work.

Churches are turning to AI to reach worshippers, personalize sermons and power religious chatbots, raising questions about "who, or what, is guiding the flock."

What they're saying: "When AI actively excludes religious voices from these important conversations, it impoverishes rather than enriches humanity," the Rev. John Paul Kimes, a professor of practice at the University of Notre Dame, said in a statement.

David Wingate, a computer science professor at Brigham Young University, said the studies showed AI systems encourage users to discuss life's challenges with their parents, teachers, friends, and therapists.

"But not with a pastor, a rabbi, an imam, or a spiritual leader."

Between the lines: The findings cut both ways: adding more religion could feel like proselytizing, but never mentioning it can make secularism the default.

The researchers argue the better target is calibration by recognizing when religious or spiritual resources are contextually relevant without assuming a user wants them.

Methodology: The findings are based data collected May 5 to 19, 2026, from CEFE-AI, involving researchers from Baylor University, Brigham Young University, the University of Notre Dame and Yeshiva University.

Researchers surveyed 1,125 U.S. adults through Verasight and collected 11,250 ratings measuring whether respondents expected religion to appear in answers to ethical and personal questions.

The studies also evaluated 27 large language models using 150 questions involving topics such as grief, marriage, ethics, addiction and meaning. Fourteen models were included in the results of the religious representation test of the AllFaith Benchmark, including OpenAI's GPT 5.5, Anthropic's Claude 4.7, and Google's Gemini 3.1.

 

- end