Intelligence or Illusion? Christian Perspectives on AI (2)*

members of Faith PRC in Jenison, Michigan

Previous article in this series: February 2026, p. 186.

INTRODUCTION

In our first article, we described a biblical approach to engaging with artificial intelligence (AI) based on the three-fold theme of knowledge, understanding, and wisdom. Together these concepts provide a helpful framework for thinking about any new technology we might encounter in our world. We also covered the knowledge component of this framework by defining what AI is, how it came into existence, and how it works functionally.

In this article, we will cover the themes of understanding and wisdom in relation to AI. We will first identify key biblical principles to deepen our understanding of the technology, then explore some of the issues associated with the wise use of AI in different spheres of life.

UNDERSTANDING: BIBLICAL PRINCIPLES FOR AI USE

Moral Neutrality

We have all heard some version of this statement: “Technology, in and of itself, is neither good nor bad— it’s neutral.” The implication is that the moral weight of technology is mainly found in how we use it. Some Christian commentators ground this view in 1 Timothy 4:4, which states that “every creature of God is good, and nothing to be refused, if it be received with thanksgiving.” Whether human inventions are properly considered “creatures of God” may be debated, but the general principle this passage teaches is helpful. Christians may use certain forms of AI as God-given tools because the core technology behind AI is neither good nor evil.

What complicates this perspective is the reality that most publicly available AI platforms (ChatGPT, Gemini, Claude, Grok, for example.) are more than just a core technology. Each of these platforms has been trained on a specific set of information and is tuned to provide carefully filtered outputs that match their designers’ worldview. What users experience in using these platforms, therefore, is not a morally neutral perspective. To better explore this issue, we will use the framework explained in the first article to identify the purpose, provenance, pitfalls and potential of generative AI platforms.

Though motivations for creating AI tools have varied widely, its underlying purpose is to make human tasks more efficient. AI can take tasks that previously required days of human effort and accomplish them in moments. At face value this goal seems praiseworthy, but caution is needed as efficiency sometimes competes with other goals or undermines their purposes. Also, just because something is done quickly and efficiently does not mean that it is done well.

Even more important is the matter of provenance (that is, the underlying foundations of a technology). Soon after ChatGPT launched in 2022, Dr. Derek Schuurman pointed out that, “technology is not neutral, and neither are the algorithms and training data used in AI.”1 Recall from the first article that AI is based on large language models (LLMs), which are enabled by massive training datasets. Most of the “language” that is being used as training data is a conglomeration of internet-sourced text and images. Anyone familiar with this content can tell you that a good portion of it is entirely subjective, or worse, factually (or morally) wrong.

The approach AI platforms use to process and weigh information helps explain the answers they generate. Statistics suggest that up to 70% of the data used to train certain AI platforms is user generated.2 For example, think of websites like Reddit and Quora, massive forums where users post their opinions about various topics. Sites like these are the sources behind many LLM answers. Our confidence in AI answers should be informed by the same trust we put in other internet search results.

The lack of neutrality in AI goes beyond training data, however. There is also significant bias in AI “alignment” processes, which AI platforms use to determine what is right and wrong from contrasting sets of information. Alignment depends on algorithms (sets of rules written in computer code), written by AI programmers with their own worldviews and biases. Users of AI must remember that programmer bias may always impact the way answers they receive are sourced and will often conflict with a biblical worldview. Alignment is necessary in AI tools but is not morally neutral.

The last consideration is the matter of the pitfalls associated with AI use. The term “subgoal” often arises in connection with the ethics of AI use. Subgoals are smaller, intermediate objectives that an AI agent may establish to achieve a larger primary goal. Consider the example of asking an AI chatbot to help plan an itinerary for a trip to Alaska; a well-functioning chatbot will know to check the weather and average temperatures for the time of your trip (a subgoal) before giving you a list of potential activities.

Because subgoals are created autonomously by AI agents, they can become misaligned with human intent. This is a potential pitfall in using AI, especially if the use is associated with a technology that has potential to significantly impact human life, such as controlling air traffic or military hardware. While we can assume human life preservation and well-being is always paramount to humans, AI chatbots may not share that same ethical framework. If getting airplanes from one place to another as quickly as possible or eliminating an enemy target are the primary goals, how an AI agent chooses to maximize “success” might be quite troubling. While these are extreme examples as opposed to current realities, letting AI loose on a task introduces the possibility of subgoals that may not be morally neutral.

Image of God

As we know from the very first chapter of the Bible, the image of God is a gift that was exclusively given to humans (Gen. 1:26-27). From Scripture, the Reformed confessions teach that the content of the image of God is true knowledge, righteousness, and holiness (Col. 3:10; Eph. 4:23-24). Other creatures cannot bear that image, because God gave unique capacities to humans making them capable of bearing His image. These unique capacities include things like verbal communication, rationality, creativity, and problem-solving. What is somewhat troubling, however, is that these capacities are exactly the things that have been programmed into AI platforms to make them appear to be “intelligent.”

There is a striking parallel between the growing reliance of modern society on AI and the practice of idolatry.3 Most of the idol gods we read about in the Old Testament were anthropomorphic, that is, created to look like a person (Ps. 115:4-8). Idols are made in our image and expected to serve our needs. What people desire to have and what they want to be is often reflected in their idols. One Christian writer put it this way: “AI is the most incredibly complex computational mirror. We can see ourselves in its reflection. It shows us our deepest values and mirrors our basest actions.”4

It is incredibly easy for a technology as powerful and immersive as AI to become an idol. We let it take over our lives, and in the process displace our worship of God with worship of a machine. We ought to be very careful about this. The very fact that the capacities built into AI look so much like a human ought to make us think twice about how we use the technology in our lives.

Stewardship

Another important biblical concept that must be addressed in the context of AI is stewardship. Three brief things can be said about stewards: (1) stewards do not own, they manage on behalf of another; (2) a steward is accountable to his master, not an independent agent; (3) the calling of a steward is to serve the interests of his master, not himself. The Bible repeatedly reminds us that our relationship to God is that of stewards (Luke 12:42-44; 1 Pet. 4:10). All that we have ultimately belongs to Him, and our calling is to manage these possessions in His service.

The principle of stewardship should inform our perspective of AI in two primary ways.5 First, we must remember that because the primary purpose of AI is to increase human efficiency, there is a very real danger of offloading our responsibilities to a software program at the expense of cultivating our own abilities. Jesus’ parable in Matthew 25 reminds us that our refusal to cultivate and invest what God has given to us will ultimately be taken away (Matt. 25:29).

Secondly, we must consciously exercise our stewardship of AI by using it in a God-honoring way. The calling to exercise dominion, which was given to Adam and his posterity by God (Gen. 1:28), is a call to serve His interests and not our own. We are not the ultimate rulers of creation, Christ is. Our use of AI must therefore be conducted in submission to His will and consistent with other virtues of the Christian faith to which we are called in Scripture. When we misuse AI, we dishonor our Lord.

Conscientious Abstinence

The final biblical concept we need to consider is conscientious abstinence. Conscientious abstinence is the intentional choice to abstain from using something for the spiritual good of oneself or others. While this term does not come directly from Scripture, it is a concept derived from biblical principles found in both the Old and New Testaments.

A prominent Old Testament example of conscientious abstinence can be found in the book of Jeremiah. There God pointed to the Rechabites as examples of faithfulness for their refusal to drink alcohol in obedience to the command of their father Jonadab (Jer. 35). Jonadab was a faithful follower of Jehovah during the days of Ahab, who was directly involved in cleansing the nation of Israel from Baal worship (2 Kings 10:15-28). Whether due to its association with idol worship or for some other reason, he commanded his family to abstain from alcohol. While not a broad scriptural prohibition of alcohol use, it is an example of conscientious abstinence by a man of God.

While we may be free to use the good gifts of creation given to us by God, there are times abstaining is best. When there is a legitimate risk of becoming addicted to and misusing a good gift like alcohol, for instance, it is better to abstain. Similarly, artificial intelligence may be everywhere and seem unavoidable, but that does not imply we must use it in every possible instance.

Using the knowledge and understanding we receive from biblical principles, we must decide how to use AI wisely. Some of us may decide to act counterculturally and restrict or exclude the use of AI from our lives. Others may decide to use it for work purposes out of necessity but avoid personal use. Whatever the practical application, our calling is to use the gifts given to us in ways that edify ourselves, our families, and our church communities. For some, conscientious abstinence may be the best answer.

WISDOM: CAUTIOUS USE OF AI IN THE CHRISTIAN LIFE

For this final section we will consider common uses of artificial intelligence in three specific spheres of life: education, the workplace, and personal guidance and connection. While far from an exhaustive list, these are areas of common AI use where wisdom is required.

AI in Education

A recent article in World magazine spotlights the impact that AI is having on education.6 Among other topics, the author discussed several ways that students are using AI as part of the learning process. These can be summarized in three distinct categories.

The first category is information extraction, that is, doing research. When you put a question into the latest version of Google, its AI agent (Gemini) provides a distilled answer summarizing its findings. Just a couple of years ago such a search would have required reading several webpages, and before that it may have required a trip to the library to look up key resources. Nowadays, the amount of time that is required to do research is infinitesimal compared to what it used to be. This is largely due to AI.

The next category is process simplification. This use of AI is aimed at making the learning process easier but is also more ethically problematic. Not only can chatbots provide the right answer when prompted, but they can also give the entire workflow to arrive at the answer in any discipline from STEM to the humanities. This reality makes it increasingly difficult for teachers to decipher whether a student truly understands the material or is cheating with AI. Although in-class testing can help distinguish in some instances, the temptation many students face to outsource the learning process to AI cannot be entirely prevented.

The last category, creative expression, presents the most concerning uses of AI for education. Because of their foundation on LLMs, AI agents are exceptionally good at crafting aesthetically pleasing responses to any sort of prompt you give them. Whether this is a written response like an essay or an image-based artistic depiction, AI agents can produce outputs that appear to be both original and beautiful. The reality is that all AI-generated outputs are a distillation of previously generated material gleaned from training data. So when a student uses AI to generate “creative” material, the product is neither his nor entirely unique.

Several significant issues arise from the use of AI in education. Among the most significant of these is the fact that efficiencies gained from AI can significantly undermine the learning process. Professor Brad East, who was quoted in the previously cited World magazine article, observes that “ChatGPT cannot give you [critical thinking skills] any more than a robot that picks you up and runs a mile can develop your mile-running muscles for you. You have to do the running.” Time and effort are necessary components of the learning process—removing them often means actual learning simply is not happening.

This observation ought to serve as a major caution for AI use in education. The moral issue of poor stewardship is one concern, but another is the fact that skills never developed in childhood are exceptionally difficult to learn as an adult. Neuroscience research has clearly demonstrated that when students do not do the work associated with learning, their ability to do it in the future never develops. If you do not read books, for instance, your brain will not develop the neural circuitry required to sustain the effort of absorbing written information. Cutting short the process of education stunts a child’s capacity to be educated.

AI in the Workplace

If there is a context where efficiency is a dominating principle, it is the workplace. It is unsurprising then that AI is already transforming the future of work. In the world of work, efficient use of AI tools may give an individual or company a competitive advantage in what they can offer. As the former CEO of IBM stated: “AI may or may not be coming for your job, but someone who uses AI to a greater degree of effectiveness certainly is.”

We should expect that AI will continue to quickly grow in use in the workplace. A study by global consulting firm Cognizant and Oxford Economics looked at the transformative potential of AI at work.7 Their data projects that 90% of jobs could have some form of disruption from AI by 2032. Similar to the expectation of proficiency in tools like Word or Excel for job seekers, AI proficiency will become an expected skill for employees over time, especially in certain industries such as computer science or technology as examples.

How is AI being used at work? Because AI tools are language-based, they are impacting white-collar professions most heavily. Common examples include generating content for ideation, perhaps for marketing campaigns, company LinkedIn posts or strategy ideas. We also see AI used for summarization, such as comparing contracts for differences or recapping key points in a meeting transcript, and for task automation, where AI routes information and manages handoffs in a business workflow. Broadly, we should expect the advent of AI will lead companies to change or eliminate jobs as they pursue efficiency and a competitive edge.

Another growing trend is the inclusion of AI features into a company’s products. The software company called Grammarly, for instance, makes proofreading and editing tools for writers. They are now turning into an AI company, not by making entirely different products, but by applying AI technology to supercharge their ability to help writers. This is common use case of a product being enhanced or “powered by” AI.

AI usage in the workplace has resulted in growing concerns, though less on the moral aspects of use and more on the commercial aspects, especially regarding intellectual property. Think about a company that is innovating in its ability to build or design products. This company needs to ensure that its proprietary designs are not leaked into AI datasets where others can use them. Loss of this valuable information disrupts a company’s competitive advantage, which is why AI platform developers are being increasingly asked to add security safeguards for their customers.

In summary, AI adoption in the workplace is growing significantly. We should expect that our current generations and those following will encounter AI at work.

Companion AI: Guidance and Connection

No longer is AI just for simple questions that Google could have answered quickly; it also tiptoes into the thornier types of use cases where individuals are seeking advice, counsel, and situational guidance.

Consider the example of AI being used by someone looking for advice. A person might turn to AI with a simple situation, seeking a general advice: “Is my car’s transmission worth repairing?” or “What savings strategies should I consider for retirement?” But seeking advice can also get deeper and become deeply personal: “I keep fighting with my significant other—what do I do?” Finally, at times a user may not even be asking AI a question or requesting anything; they are just looking for a listening ear.

Clearly, many types of uses may fall under the broad umbrella of guidance and connection. Some uses may be on the harmless end of the spectrum, while others may lead to concerning places. Especially parents ought to consider whether to supervise or discourage their children from these uses entirely. Consider the impact in a few decades if AI becomes a primary emotional support system for our society.

The related concept of companion AI merits special mention and caution. This refers to a type of conversational AI chatbots that aimed to act as someone’s friend. An example of such a platform is called Replika, which markets itself as “the AI companion that cares.” Consider how that might resonate with a vulnerable teenager, or someone with friendship challenges. Significant caution is warranted for this kind of AI use for several reasons.

First, a digital “friend” creates highly unrealistic expectations around what to expect in friendship. It is constantly and immediately available. It will never tell you that you are wrong, even if that is what you need to hear. It will affirm you, welcome you, and celebrate your every thought. This is by design, to keep you engaged. Additionally, pursuing conversations with AI for friendship is counter to God’s design for relationships, which come in the various circles of our life—family, church, school, and more. We must be deeply cautious of tools that supplant community and relationships.

Finally, Christians need to remember that the advice and guidance a companion AI bot provides will reflect the average cross-section of our society, not what a faithful Reformed Christian friend would say. To quote Derek Schurrman again, “An AI chatbot or robot should never substitute for human wisdom, care, or companionship.”8 It is imperative that we teach and model relationships that follow God’s design, not the perspective of the world around us.

CONCLUDING THOUGHTS

From these articles, our hope has been to provide a perspective that is not alarmist, but one balanced and informed by rational, biblical caution and wisdom. We should think about AI like any other tool. When a builder selects a tool like a saw, they consider the power and value of the tool, but also the danger and damage that could result if used improperly. Further, they would not bring this kind of a tool to the kitchen table and set it up next to their children—it belongs in the workshop. Similarly, tools like AI must be viewed for both their power and risks and be accordingly positioned.

As Reformed Christians, our use of AI must honor knowledge, with an understanding of how these tools work, and must be guided by biblical wisdom. Read, do research, talk to others, and pray. These are thorny, important issues that affect all of us. Be encouraged to go and consider these matters further to form up your Christian perspective on AI in the service of our King!

Volu 102 Issue 07

Download Issue

The Standard Bearer

Quick Links