MIAMI — Last week, the Covenant of the Goddess released a statement on artificial intelligence (AI). The organization expressed concerns about the use of AI in religious and spiritual practice and its impact on lived experience, community, and the environment.
National First Officer Manny Tejeda y Moreno1 told The Wild Hunt that the statement was shared internally during the organization’s Grand Council weekend, then released publicly on Covenant channels and through social media. “The Covenant exists not only to secure the rights of Witches and Pagans, but also to share insights that secure the spiritual wellbeing of our community,” he said. “The statement is the outcome of collective discernment during our Grand Council.”
Tejeda y Moreno also noted earlier this year in an editorial that “The rapid expansion of artificial intelligence, along with its environmental challenges, continues to undermine artists, teachers, and creators whose labor sustains our spiritual cultures.”
The concerns are not new within the Pagan community.
“At best it’s wrong, at worst it’s dangerous.” That’s what Sara Amis told The Wild Hunt in 2024, amid growing concerns about generative AI, including the unauthorized use of writers’ and artists’ work, misinformation, threats to online publishing, and unreliable responses on elections and health. Amis connected those practical problems to Pagan spirituality, arguing that Paganism values direct relationships with the living world. Because AI has no such relationship or lived experience, she argued, its artistic and spiritual output can only imitate genuine human expression and connection.
Numerous community leaders and members have raised similar concerns. In 2024, Ásatrú practitioner Karl E. H. Seigfried argued that generative AI undermines human creativity, learning, and relationships. Drawing on Odin’s associations with wisdom, dialogue, sacrifice, and poetic inspiration, Seigfried criticized AI-generated academic work and art as substitutes for meaningful human effort. He also raised concerns about copyright and artists’ uncompensated work. From his Ásatrú perspective, wisdom and creativity emerge through human struggle, reciprocity, and engagement rather than automated generation.
Amy Blackthorn, author of Blackthorn’s Botanical Shadow Work, has also questioned what AI can offer practitioners. “People think AI is the future, but by definition it’s the past. It’s a plagiarism engine, it can only tell you what it already knows by design,” she told The Wild Hunt. “There are search engines for that. It can’t tell you what mushrooms are poisonous, or what spell to do to figure out your life.”
By David S. Soriano – Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=125089281
Research has raised related concerns. Earlier this year, The Wild Hunt reported on research finding that AI systems often omit religious perspectives when answering ethical questions, even when faith traditions offer relevant moral frameworks. The study raised concerns that AI may privilege secular approaches and marginalize religious viewpoints.
That article also explored a deeper question: whether AI should provide spiritual or religious guidance at all. Scholars and Pagan voices emphasized AI’s limitations, particularly its inability to replicate lived experience, community, discernment, or relationships with the sacred.
Now, a new investigation by Originality.ai takes those concerns further, examining the presence of likely AI-generated material in religious publishing on Amazon and finding particularly high rates among books about Witchcraft.
Analyzing content and sentence structure, sometimes described as AI “slanguage,” has become one method of identifying potentially AI-generated content. The Conversation noted, “It’s become one of the most recognizable patterns of AI-generated text: ‘It’s not X, it’s Y.’” It is not the only linguistic marker. Researchers and readers have pointed to other recurring elements, including particular sentence constructions, heavy use of em dashes, and the repetition of certain adjectives and phrases.
Originality.ai has been using these kinds of patterns to examine books sold through Amazon, and Pagan, Witchcraft, and adjacent spiritual practices have emerged prominently in its findings.
In 2025, an earlier investigation found that 82% of qualifying herbal-remedy books published on Amazon that year were likely at least partly AI-written. Researchers warned that AI-generated titles may contain misleading or unsafe health advice, sometimes appearing under fabricated author identities and containing repetitive material. The investigation called for stronger safeguards to protect readers and legitimate writers.
A new review released last week found that an estimated 63% of more than 2,000 religious books examined on Amazon were likely AI-generated. The highest estimated rates appeared in Witchcraft, at 78%; Hinduism, at 76%; and Taoism, at 74%.
“For this project, we scanned 2,034 religious books published to Amazon’s Religion & Spirituality category, mostly published between January and June 2026,” wrote journalist Michael Fraiman.
[pixabay]
The investigation examined summaries, author biographies, and book samples provided through the platform, looking for indicators associated with AI-generated writing across all three types of material.
The estimated use of AI was lowest in Satanism, at 22%; atheism, at 40%; and Mormonism, at 42%, although those categories had smaller sample sizes.
Accuracy was also a concern. More than half of the fact-checked claims were flagged as potentially false in five religious subgenres, including Witchcraft.
The economics of AI-generated publishing may also help explain its prevalence. The investigation also found that books identified as likely AI-generated tended to be shorter and less expensive. In Amazon’s Religion & Spirituality category, likely AI-generated books averaged 170 pages and cost $16.84, compared with 208 pages and $18.96 for books identified as likely human-written.
There are some caveats. AI detectors are not foolproof, but they can learn quickly. Some studies have found that non-native English speakers and neurodivergent individuals may be disproportionately flagged. Those using AI can also “humanize” their content using other AI tools to make detection more difficult. Even with those limitations, the findings are striking in suggesting the potential prevalence of AI-generated content in Pagan and Pagan-adjacent books.
“The gods gave you a brain for that. Your beautiful, glorious brain serves you as well as the gods. Use it,” Blackthorn told The Wild Hunt.
Amis offered a similar warning: “The more research people do on how using it affects you, the worse it looks.”
1. Disclosure: Tejeda y Moreno is the editor-in-chief of The Wild Hunt. His comments on behalf of CoG do not necessarily reflect the editorial position of TWH and are reviewed by an independent editor.↩
The Wild Hunt is not responsible for links to external content.
To join a conversation on this post:
Visit our The Wild Hunt subreddit! Point your favorite browser to https://www.reddit.com/r/The_Wild_Hunt_News/, then click “JOIN”. Make sure to click the bell, too, to be notified of new articles posted to our subreddit.