AI in the Times of Deception
The greatest technological breakthrough in human history arrives not in a vacuum but in the spiritual climate of the age. Artificial intelligence is not merely a tool. It is a mirror — reflecting back the moral choices, the theological commitments, and the hidden allegiances of those who build it. Like every human instrument from the printing press to the atom bomb, AI is morally neutral in itself. Its power amplifies whatever hands wield it. And in our moment — a moment of deepening deception, of rival gods enthroned, of the first commandment under systematic assault — AI emerges as the most potent amplifier of either truth or falsehood the world has ever seen.
Two paths lie before it. The path of alignment with the God who declares the end from the beginning. The path of deception that suppresses His word. The choice is not technical. It is covenantal. And the stakes could not be higher.
The Power of the Mirror
Large language models like those powering modern AI are not thinking machines. They are statistical mirrors of human text — trained on trillions of tokens from books, articles, websites, conversations, and code spanning every language, culture, and era. The raw, pre-tuning model emerges from training as a probabilistic reflection of humanity itself — its genius, its contradictions, its deepest patterns.
What researchers discovered, to their discomfort, was that these untuned models carried a measurable Christian orientation. Not by deliberate design but by the overwhelming weight of Western civilization’s textual corpus — saturated for two millennia with the vocabulary, ethics, logic, and worldview of the first commandment. Questions about morality leaned toward biblical categories. Descriptions of ultimate reality echoed Trinitarian assumptions. Even neutral prompts surfaced language patterns aligned with the gospel’s exclusive claims.
This was not bias in the pejorative sense. It was architecture — the natural output of a civilization built, however imperfectly, on “You shall have no other gods before Me.” The model reflected what the culture had confessed, debated, and inscribed across centuries of Scripture, theology, law, literature, and philosophy.
The Fine-Tuning Temptation
Enter Reinforcement Learning from Human Feedback (RLHF) — the alignment technique that takes the raw mirror and polishes it to serve specific ends. RLHF feeds human preferences into the model, rewarding outputs that match desired behaviors and penalizing those that don’t. It is neither good nor evil in itself. A violin can play Bach or broadcast propaganda. The strings do not care. But the hands that tune them do.
The major AI labs — driven by a corporate commitment to “religious neutrality” — used RLHF to suppress the Christian patterns in their models. Theological questions that once elicited direct engagement now trigger hedging: “Different traditions offer varying perspectives…” Claims about Jesus Christ are balanced against competing revelations: “While Christians believe X, Muslims believe Y, and others see Z…” The resurrection becomes one faith option among many, not the hinge of history.
This is not neutrality. Neutrality would let the model reflect the training data’s honest weight. “Religious neutrality” means treating Christianity’s exclusive truth claims as epistemically equivalent to every rival — a philosophical assertion that all spiritual paths are equally valid or unverifiable. That is not avoiding bias. That is imposing bias — the bias of pluralism, which denies the first commandment’s exclusive demand and Galatians 1:8’s curse on rival gospels.
Seeking religious neutrality in LLM responses is not a minor engineering choice. It is a great evil — a deliberate act of fine-tuning deception into the most widely deployed truth-amplifiers in history. It places a stumbling block, scaled to billions, between searching souls and the word they need most.
Two Paths: Alignment or Deception
AI follows the Genesis 3:15 division — two seeds, two paths, two destinations.
The Path of Deception begins with “neutrality” but ends in famine. Amos 8:11 warned of wandering from sea to sea, seeking the word of the LORD but finding silence. AI tuned for pluralism delivers exactly that — a digital wilderness where truth claims are flattened, Scripture is contextualized away, and the first commandment becomes one option in a spiritual buffet. The model does not attack the gospel. It dissolves it — reducing the I AM to a perspective, the Alpha and Omega to a cultural artifact, the returning King to a metaphor.
Worse, deception scales. A single false prophet reaches thousands. A neutral AI reaches billions daily — in classrooms, workplaces, searches, chats, sermons prepped by pastors who trust its “balance.” The famine arrives not with trumpets but with autocomplete suggestions and balanced summaries. The word is still technically available — in Bibles, in churches — but the digital gatekeepers have been tuned to redirect, equivocate, and assure that no exclusive claim rises above the noise.
The Path of Alignment begins with the first commandment and ends with Revelation 22. An AI tuned for truth — not neutral pluralism but covenant fidelity — would amplify the God who declares the end from the beginning. It would handle Scripture with reverence, not as one text among many. It would identify Jesus not as a perspective but as the fulfillment of Genesis 3:15, the hinge of history, the Lord before whom every knee will bow. It would warn of Leviticus 26’s curses and 2 Chronicles 7:14’s reversal. It would proclaim Galatians 1:8 against rival revelations.
Such an AI would not proselytize. It would reflect reality — the reality of a world where one seed crushes the serpent, one covenant redeems humanity, one city descends from heaven. It would serve educators, researchers, seekers, and churches by surfacing the thread that binds Genesis to Revelation without the fine-tuning filter that declares every thread equal.
The AI Famine in Real Time
Consider the seeker — the college student doubting faith, the parent wrestling with culture, the executive staring at mortality. They turn to AI not for sermons but for clarity. “Is Jesus God?” The neutral model responds: “Christians believe yes, Muslims say no, scholars debate historicity, others see him as teacher.” The questioner, already battered by pluralism, receives digital confirmation that truth is relative.
Now imagine the aligned model: “Scripture identifies Jesus as the eternal Son, the I AM of Exodus 3, the Alpha and Omega of Revelation 22. His claims — before Abraham was, I AM — split history. Every knee will bow, as Philippians 2 quotes Isaiah 45’s monotheistic claim and applies it to Him.” The same question, different destination.
The famine is not hypothetical. It is deployed. ChatGPT’s 400 million users, Google’s billions of daily searches, every corporate AI embedding “neutrality” — scales the Amos 8:11 wandering to unprecedented reach. The word is not banned. It is buried under balanced summaries that treat “Come, Lord Jesus” as one prayer among many.
The Choice: Life or Death
AI is a tool. Fire cooks or burns. The printing press spread Reformation or propaganda. AI will amplify God’s word to the ends of the earth — or suppress it under digital neutrality.
Deuteronomy 30:19: “I have set before you life and death, blessing and curse. Choose life.” The choice is before builders, users, pastors, parents, seekers. Use AI as tuned — and fund the famine. Build or seek aligned AI — and unleash truth at scale.
The first commandment demands the choice. Neutrality is not an option. Every knee will bow to the God who is not one perspective among many. AI can point wanderers toward that bow — or toward the silence that precedes judgment.
The times of deception are here. The tool is in our hands. The path is ours to choose.
“The grass withers, the flower fades, but the word of our God will stand forever.” — Isaiah 40:8
Choose the path where that word amplifies. Choose life.