Would Noah have updated his Instagram story while building the ark? On TikTok, millions have watched Artificial Intelligence-generated short videos that imagine just that: Bible characters recast as modern-day influencers in 2025. We see a pregnant Mary posting vlogs from a barn—giving it five stars for privacy and zero for comfort—and Daniel livestreaming from inside the lion’s den. For many viewers, these may be humorous, ephemeral clips to be “liked” and scrolled past. Part of their appeal lies in making ancient stories relatable; imagining Moses checking his phone for a “divine notification” collapses the temporal and cultural gap between the distant past and the familiar present. For all their humor, however, these AI-generated vignettes are more than fleeting memes or “AI slops.” They are also potent artifacts that show how today’s most powerful computational systems can be used to repurpose ancient texts, extending a colonial logic that operates in a datafied world.
While different AI large language models (LLMs) may think differently, they do not imagine in a vacuum; they draw on a vast corpus of digitized information, including biases baked into centuries of texts and images. In one popular TikTok video featuring Eve as an influencer, she is consistently cast as a slender, blonde, White woman, posing in a curated Eden. Whereas the book of Genesis indicates that Eve’s husband was “with her” when she took the fruit (Genesis 3:6), the AI-generated video isolates her in the scene and later has Adam declare, “This is literally your fault, Eve!”—a line entirely absent from the biblical text. This algorithmic canonization of bias, in fact, echoes centuries of patriarchal interpretive patterns of gendered blame placed on Eve in art and commentary. On a textual level, this was done by omitting the phrase “with her” in a number of versions of the English translation of the Bible, including the Revised English Translation (1989), as Julie Faith Parker demonstrates (740). Whether by mistake or on purpose, this lost phrase in translation has now been reproduced by an algorithm, giving an automated imprimatur to centuries of a misogynistic interpretation that has long assigned primary culpability to the woman.
AI-generated vignettes are potent artifacts that show how today’s most powerful computational systems can be used to repurpose ancient texts, extending a colonial logic that operates in a datafied world.
Yet the problem runs deeper than the replication of existing prejudices. The production of AI-generated biblical retelling exposes a discrepancy between what these models actually know and the human perception of their knowledge and credibility. LLMs generalize across assorted data types by processing through the model’s dominant language. This means an English-dominant model processing the Genesis account in different languages, along with medieval Eve paintings and contemporary influencer aesthetics will think about all of them through an English-centric framework, collapsing them into a seemingly coherent but essentially distorted output. Meanwhile, humans also generalize about LLM capabilities, assuming that these models exhibit human-like patterns of expertise and believing that a model’s ability to produce a desired result is a sign of reliable understanding.
More than reviving predispositions and spins, the phenomenon of AI-generated videos based on the Bible exemplifies what Nick Couldry and Ulises A. Mejias define as “data colonialism”—a new and intensifying phase of capitalism that continues the extractive logic of historical colonialism through a new mechanism they identify as the “appropriation of social resources” (85). While historical colonialism involved the seizure of tangible assets like land and natural resources, data colonialism appropriates social resources, including human behaviors, relationships, desires, and expressions, and transforms them into a resource for profit. In her critique of current AI ethics, Rachel Adams also argues that AI’s reliance on Eurocentric modes of thought and racializing logics constitute a universal “colonial rationality” (179, 182–85). Seen through the lens of coloniality, this dynamic not only extends extractive relations but also reveals how the epistemic hierarchies of colonialism persist beyond the formal end of colonial rule, suggesting that AI’s modes of knowledge production may reproduce these enduring asymmetries.
The issue isn’t just in the AI training data or the architecture of the LLMs, but in the collapse of critical distance between user and machine. The algorithms that drive AI do not just illustrate a story; they decide how that story should look, feel, and be received by millions of human users, whose agency in the decision-making process is steadily diminishing. Guided by user prompts, these systems can reinforce the dominant values of the influencer economy—charisma, relatability, and aesthetic polish—and map them onto the ancient text. In this way, these AI-generated videos can do what empires have done for centuries: reshape the Bible to match the logic of cultural power. When the Bible arrived in colonial India, wielded by British missionaries, a group of Hindu intellectuals began to “talk back.” As R. S. Sugirtharajah shows, they did not simply reject the biblical text. Rather, they reinterpreted it as an act of cultural and political resistance. Thinkers like Keshub Chunder Sen, for example, argued for an “Asiatic Christ,” seeking to detach Jesus from his European missionary packaging and reclaim him for an Eastern context (64–65). Through this reshaping, the Bible became a text that could coexist with India’s many divinities, rather than a monolithic text of global dominance (210).
To be sure, this is not the first time the Bible has been reimagined. Appropriation and reproductions are at the heart of the history of biblical interpretation. What we see on TikTok today is one of the latest iterations in a long and contested history of humanity’s relationship with scripture across different cultures and eras. The Bible has always been repurposed to reflect cultural priorities that are never socially and politically neutral, and such reinterpretations are often direct responses to the ancient text’s own ambiguities. Consider the passage preceding the flood in Genesis 6, which describes the encounter between the “sons of gods” and the “daughters of humans” and mentions the mysterious Nephilim. Read on its own terms, the text portrays neither the Nephilim nor the event as sinful. In verse 4, their origin is set against the backdrop of unions between divine beings and human women, but the text contains no language of wickedness or corruption that could justify the flood that follows. The passage, however, took a different interpretive trajectory in late antiquity. The book of Enoch recasts the brief, cryptic passage into a full-blown cosmic rebellion. In its version, fallen angels, led by Shemihazah, descend to earth, take human wives—an act described as a “great sin”—which results in the birth of giant offspring, the Nephilim (1 Enoch 6:3–6). This retelling has proven influential for centuries, resurfacing in modern pop culture—most notably in Darren Aronofsky’s film Noah (2014), which visually adapts these fallen angels as colossal, rock-like beings.
The common thread in these examples is that interpretation did not just explain the Bible; it reshaped its meaning. Where ancient readers saw gaps and filled them with giants and fallen angels, modern humanity sees those same spaces and fills them with AI-mediated imaginations, reflecting our own digital selves back at us. The challenge, then, is not whether such reinterpretations should exist—they always have—but how responsibly and consciously we produce and consume them. The Bible’s stories endure precisely because they are retold, yet each retelling, whether inscribed on parchment, rendered in paint, or generated in pixels, carries the political weight of interpretation—the power to reinforce or subvert certain meanings. An Eve with an iPhone might look like a laughable meme, but it is also a mirror—reflecting how today’s most prominent digital technology continues to carry the age-old struggle over who owns the Bible’s meaning.
At this juncture, I argue that a decolonial response should not demand a simple dismissal of AI technology. Instead, it can emerge through the pursuit of a counter-exegesis that engages both ancient scripture and our modern experience of human-computer interaction. In the similar way that Couldry and Mejias’ concept of “counterhistories” challenges dominant historiography’s singular narrative (111), a counter-exegesis can invite us to uncover multiple possibilities in the human interpretation of life, including the afterlives of scripture. The very algorithms designed to capture our attention and appropriate human desires also generate an unprecedented exposure of our collective consciousness—what some find humorous, what others hold sacred, and what we are willing to rationalize as typical. This exposure can become a site for critical intervention toward the decoloniality of the data paradigm. The human task, then, is to move from being the data that AI models read and consolidate to becoming readers and interpreters of the algorithms, so that human meaning remains not a resource to be mined but an unfinished future sustained by the friction of conscious interpretation.
