Educators are confronting a new reality as AI tools like ChatGPT become widespread among students. Traditional take-home assignments and essays are increasingly at risk as students commonly use AI chatbots to complete schoolwork.

Schools are responding by moving more writing tasks into the classroom and monitoring student activity. Teachers are also integrating AI into lessons, teaching students how to use it responsibly for research, summarising readings, or improving drafts, rather than as a shortcut to cheat.

Policies on AI use still vary widely. Some classrooms allow AI tools for grammar checks or study aids, while others enforce strict bans. Teachers are shifting away from take-home essays, adopting in-class tests, lockdown browsers, or flipped classrooms to manage AI’s impact better. 

The inconsistency often leaves students unsure about acceptable use and challenges educators to uphold academic integrity.

Institutions like the University of California, Berkeley, and Carnegie Mellon have implemented policies promoting ‘AI literacy,’ explaining when and how AI can be used, and adjusting assessments to prevent misuse.

As AI continues improving, educators seek a balance between embracing technology’s potential and safeguarding academic standards. Teachers emphasise guidance, structured use, and supervision to ensure AI supports learning rather than undermining it.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Disclaimer and Acknowledgements: Answers in Genesis and the author are grateful to the various pastors who reviewed this paper and provided input. This paper cites a variety of scholarly sources that may not reflect the views of the author or of Answers in Genesis.

Introduction

A new era has dawned in church history. For over two millennia, Christians who gathered to hear teachings from God’s Word could safely assume the teacher would be human. In June 2023, however, over 300 people flocked to St. Paul’s church in Fuerth, Germany, to hear from a virtual preacher powered by artificial intelligence (AI).1 The following year, Christian headlines reported, “Pastor Creates AI Version of Himself Offering Personalized 1-on-1 Prayer.”2 Similarly, websites such as ai-pastor.com advertise individually tailored sermons, devotionals, Scripture analyses, and prayer support from chatbot “pastors.”3 Not even having a flesh-and-blood pastor guarantees hearing fully human-authored sermons, given how companies like SermonDone encourage pastors to copy, paste, and preach AI-generated messages.4

With AI unlocking novel possibilities and raising new questions within ministerial contexts, pastors need biblical, ethical boundaries for using AI in ministry. In response, this four-part paper argues that, while AI can be a useful support for certain purposes such as research,5 AI should not replace pastors’ spiritual leadership responsibilities such as personal Scripture study, sermon preparation, and pastoral care.

Part One examines a biblical understanding of four concepts: the church, the pastor, the sermon, and technology. Part Two introduces the basics of what generative AI is, how this technology works, and how AI differs from human intelligence. Part Three describes common practices and rationale regarding the use of AI in pastoral ministry. Finally, Part Four applies theological concepts from Part One to argue against outsourcing personal Scripture study, sermon preparation, and pastoral care to AI.

Part One: Toward a Biblical Theological Understanding of Pastoral Ministry

The Ministerial Context: The Church

To draw biblical boundaries for AI in churches, we need to consider what God’s Word says about the nature, purpose, and structure of the church. We can glean initial insights into the church’s nature from considering the word church itself, known in the Greek New Testament (NT) as ἐκκλησία (ekklesia).6 The Septuagint renders ekklesia as the Hebrew term קְהַ֖ל (qahal), which exclusively refers to the physical assembly of God’s covenant people.7^, 8 The NT use of ekklesia also refers to an embodied community of God’s people, now under the new covenant in Christ.9^, 10 Scripture depicts this new covenant community as God’s flock (1 Peter 5:2) and as Christ’s own body and beautiful Bride (Ephesians 5:29–32). Ultimately, Revelation 7:9 portrays the church’s fulfillment as an eschatological gathering of all God’s redeemed people.11 By nature, then, the church is an embodied, eschatological assembly of human believers unified in Christ.

The church’s purpose and structure reflect its nature.

The church’s purpose and structure reflect its nature. On earth, the church gathers in local congregations that function as “embassies” of God’s kingdom, pointing to heavenly realities.12 In view of these realities, the church works to make disciples (Matthew 28:18–19) and to edify, equip, and mobilize believers as Ephesians 4:11–16 describes. This passage lists various offices of church leaders, including “shepherds and teachers,” whom God ordains to work together for “building up the body of Christ” (Ephesians 4:11–12). The church’s structure thus flows from its God-given purpose, with human leaders serving God’s people under Christ, the church’s head (Ephesians 5:23).

The Ministerial Person: The Pastor

As 1 Peter 5:2 indicates, the pastor’s high calling is to serve as an embodied, human shepherd of God’s embodied, human flock (cf. John 21:15–17).13 How does Scripture portray the shepherd’s responsibilities? Primarily, the shepherd feeds the sheep by preaching God’s Word (1 Timothy 4:13; 2 Timothy 2:15, 4:1–2). The shepherd also leads the sheep by personal example, following the example of Christ (1 Timothy 4:12; 1 Peter 5:3; cf. 1 Corinthians 1:11). Meanwhile, the shepherd protects the sheep by defending against false doctrine (2 Timothy 4:3). The shepherd additionally corrects the sheep who are going astray, guiding them through appropriate exhortation, reproof, and discipline (2 Timothy 4:2). Along the way, the shepherd tends to sheep in need by providing pastoral care.14 In all these responsibilities, the pastor answers to the chief Shepherd, Jesus (1 Peter 5:4).

The habits, hallmarks, and heart of the shepherd harmonize with these responsibilities. The habits of the shepherd include deep personal Scripture study and meditation, prayer, accountability to others, and the active pursuit of godliness (1 Timothy 4:7–10, 6:11; 2 Timothy 2:21–23). The hallmarks of the shepherd include personal diligence (1 Timothy 4:7–16; 2 Timothy 2:15), integrity (1 Timothy 4:7–16, 6:11; 2 Timothy 2:21–23; Titus 1:5–9), perseverance (2 Timothy 2:1–6), and faithfulness, resisting compromise in both lifestyle and teaching (1 Timothy 6:13–14). And the heart of the shepherd beats for something higher than itself. The pastor is not to be driven by self-centric, earthly focused motivations (1 Peter 5:2) but rather by love for the awesome chief Shepherd (1 Peter 5:4).

The Ministerial Message: The Sermon

Along with characterizing shepherds as teachers in Ephesians 4:11, Paul’s letters to church leaders consistently emphasize the importance of teaching sound doctrine (e.g., 1 Timothy 4:13; 2 Timothy 2:15, 4:1–2; Titus 2:1). This emphasis highlights the role of preaching as a primary responsibility for pastors. In Scripture’s portrayal of preaching across both the Old Testament (OT) and NT, God’s human messenger communicates God’s Word to God’s people in a particular time and place. Nehemiah 8:8 depicts preaching as both proclaiming and explaining Scripture, a pattern that carries into the NT in 1 Timothy 4:13.15 The scriptural pattern of preaching is God-centered, Bible-focused,16 and aligned with the biblical responsibilities, habits, hallmarks, and heart of the shepherd.

Scripture also clarifies the purpose of preaching. As theologian David Christensen surmises from Colossians 1:28, “Our purpose in proclaiming Christ is nothing short of leading each person to completion in Christ.”17 In line with this purpose, preaching is context specific. The sermon delivers God’s Word to this group of God’s people in this place at this time with these needs.18^, 19 Discerning and communicating God’s heart for this congregation requires diligent prayer, study, reflection, and relational reliance on Christ.20^, 21 As Paul David Tripp said so well,

Preaching is more than the regurgitation of your favorite exegetical commentary. . . . It is bringing the transforming truths of the gospel of Jesus Christ from a passage that has been properly understood, cogently and practically applied, and delivered with the engaging tenderness and passion of a person who has been broken and restored by the very truths he stands up to communicate. You simply cannot do this without proper preparation, meditation, confession, and worship.22

Underscoring this need for personal diligence, Scripture repeatedly characterizes preaching as labor (1 Timothy 5:7). Christensen unpacks the NT definition of labor as involving struggle, striving, and even a sort of agony, adding that God’s Spirit empowers this labor.23 Part of this labor requires careful study,24 along with basking in the richness of Scripture as only a human in relationship with God can do (Psalm 119:18).25

The Ministerial Media: Technology

Whether in the form of paper and ink or pixels and screens, technologies can support (but cannot appropriately supplant) human ministry. As software engineer and theologian John Dyer describes, technology is a gift to steward for the good of humans and creation.26 Dyer also stresses that no technology is value neutral. Instead, every technology reflects the values of its users and developers. Additionally, even well-intended uses of technology can yield non-neutral consequences. For these reasons, wisely stewarding technology requires anticipating how specific applications of technology may impact our thinking, behaviors, and relationships.

Whether in the form of paper and ink or pixels and screens, technologies can support (but cannot appropriately supplant) human ministry.

Potential impacts on relationships are especially significant because God designed humans as intensely relational beings, fashioning us for fellowship with himself and others. Technology cannot substitute for these relationships. Nor can technology fulfill the relationship-based vocations and responsibilities to which God has called humans, including loving others, parenting, and discipling. For all these reasons, Christians must proactively attend to how specific uses of technologies (even for noble motivations) may cause unintended impacts on the pastor, the church, and the ministry.

Part Two: Understanding Generative Artificial Intelligence

Turning to the technology itself, we can survey the basics of what generative AI is, how it works, and how AI differs from human intelligence. AI defies a single consensus definition, partially because researchers don’t agree on how to define intelligence. But the Merriam-Webster dictionary offers a starting point, saying AI can refer to “the capability of computer systems to imitate intelligent human behavior,” to programs that have this ability, or to the “branch of computer science” that develops them.27

Generative AI systems consist of artificial neural networks, layers of interconnecting software constructs known as computer nodes.28 These nodes function like artificial brain cells by receiving, processing, and transmitting data, allowing the program to “learn” by making its own connections based on data it receives. By analyzing vast quantities of data, the AI system gains knowledge for generating novel materials in response to prompts.

While AI’s outputs are often impressive, they’re not always reliable or accurate. AI often reflects biases due to training data or presents false data as factual.29 AI’s decision-making processes are also usually unclear, making AI a “black box” largely opaque to human scrutiny.30 Even when developers add layers of transparency to AI systems, humans tend to uncritically trust AI’s outputs, a tendency known as automation bias.31^, 32

Another vital point to keep in mind is the difference between artificial “intelligence” and human intelligence.33 Namely, AI is not an image bearer of God. It is not a living, self-aware, spiritual being. It does not consciously understand its own actions and responses. It cannot spiritually comprehend God’s Word. It cannot relate personally to God or humans. It cannot experience emotions, despite using emotional language. It cannot truly care for others. It cannot worship. And it cannot love.

Part Three: Common Practices and Rationale Regarding AI in the Church

With this groundwork established, we can examine churches’ current uses of AI. Two broad (and sometimes overlapping) categories of AI usage include content-producing tasks and non-content-producing tasks. The former tasks involve using AI to generate or manipulate text, visuals, or other materials meant for another person’s consumption, while the latter tasks do not.

Content-Producing Tasks

How are real churches applying AI to content-producing tasks?34 Some ministers report using AI for personal correspondence to generate emails, text messages, or reference letters.35 AI software marketed to pastors also allows for automating every stage of sermon preparation, which include researching, structuring, writing, editing, and titling the message.36 Other churches apply AI to marketing purposes, such as creating ads for church programs or events, or to generate images and text for social media.37 Some churches also utilize AI for public engagement (for instance, by adding chatbots to church websites) or for congregational engagement (such as generating sermon summaries and study guides for congregants).38 Pastors have also reported using AI for small group facilitation, for example by generating discussion questions for Bible studies.39 Worship facilitation represents another up-and-coming use with software marketed to worship leaders advertising, “Create powerful worship songs with an AI gospel music generator.”40 A pastor in 2023 even reported transparently hosting a fully AI-derived church service, with an AI-generated call to worship, children’s message, sermon, communion liturgy, and song.41 Attempts to automate the discipleship process itself are evident in websites like ai-pastor.com, which promises users 24/7 access to a personalized chatbot “pastor.”42

Three questions can help to further classify content-producing tasks. First, is the AI “feeding the sheep” by producing theologically instructive materials for the congregation’s consumption?43 For instance, AI that generates a sermon point from a Scripture passage contributes more directly to the congregation’s spiritual formation than does AI that generates an image for visually emphasizing a point that the pastor has already selected. Second, is the AI taking over an activity that is normally assumed to involve genuine human interaction, such as personal correspondence? Third, is the AI being applied as though it can do something that machines cannot in fact effectively, appropriately, or genuinely do, including praying, worshipping, caring, or loving? All three questions can help to gauge the (in)appropriateness of specific AI usages.

Non-Content-Producing Tasks

In contrast, what are some non-content-producing tasks? In one study, a pastor reported querying AI to research local demographics to ascertain more effective ways to reach the community.44 (Presumably, however, personally engaging with community members and asking them questions remains the surest strategy for learning about local needs.) The same study also described a pastor using ChatGPT for calendar scheduling purposes. As another example, pastors might listen to AI programs read aloud the text of a human-authored sermon, blog post, or other written materials to catch typos.

Pastors must beware of AI’s tendency to give fabricated answers.

Pastors could also use AI to identify relevant resources for further study, for instance by asking AI for examples of scholarly works on a given topic from a certain theological perspective. Relatedly, AI programs may help pastors locate specific data within lengthy academic works. For example, a pastor could ask AI whether a given report about evangelicals’ beliefs discussed the prevalence of a certain theological misconception, and if so, which page of the report contains this information. Even in these cases, however, pastors must beware of AI’s tendency to give fabricated answers, requiring diligence to verify AI’s claims.

Identifying and Responding to Five Common Arguments for Pastoral Uses of AI

Typically, advocates for the use of AI in ministry, including for content-producing tasks, appeal to at least one of five arguments. First, using AI for content production saves pastors time for other activities such as “community engagement.”45 Second, using AI prevents churches from “getting left behind” in a digitally driven culture.46 Third, AI can help churches relate better to younger, tech-savvy audiences.47 Fourth, using AI for purposes including sermon generation is just the next stage in the church’s adaptation to changing technologies, like the printing press.48 And fifth, using AI for sermon generation is “no different than consulting commentaries, books, or online resources.”49

In response to the first argument about saving time, the assumption that community engagement activities are more important for pastors than preaching lacks scriptural support. As Part One discussed, the pastoral Epistles emphasize teaching God’s Word as a primary responsibility for the shepherds of God’s flock. While other community activities are important, their significance does not biblically justify forgoing essential pastoral responsibilities (much less, passing off AI-generated content as one’s own). Christensen notes that, although a popular message implies discipleship happens mainly in relational settings outside of preaching, “the divorce between disciple-making and preaching is unbiblical.”50 Correspondingly, in Acts 6:1–6, the apostles delegated community engagement to others in order to have more time for preaching, the opposite of what advocates for AI-generated sermons argue.

Here, a caveat is in order. Pastors may want to save time from sermon preparation if they feel overwhelmed for other reasons, such as personal struggles. Certainly, more supports are needed for pastors in these situations.51 Yet given the established theological and ecclesiological significance of preaching, the solution for wounded shepherds should not be to outsource the feeding of their flocks to a robot.

Turning to the second argument, the idea that churches should prioritize cultural relevance harbors problems. This argument implicitly assumes that keeping up with the latest cultural trends is inherently good and that the church should replicate the culture. While 1 Corinthians 9:22–23 suggests that appropriate contexts exist for becoming “all things to all people,” Christians are not to follow the culture to the point of adopting unbiblical behaviors or beliefs. The usefulness of being “all things to all people” doesn’t justify boarding cultural bandwagons in ways that run contrary to Scripture (including Scripture’s emphasis on pastors’ personal teaching responsibilities). So arguments from cultural relevance cannot justify pastors using AI in ways that don’t accord with scriptural principles.

Regarding the third argument, AI isn’t necessarily the most helpful way for churches to minister to young people. Younger generations already exhibit problematic beliefs about “relating” to AI entities. For example, the Institute for Family Studies reported in 2024 that “1 in 4 young adults believe AI partners could replace real-life romance.”52 Further research suggests that, especially for young people, digital communication undermined individuals’ well-being during the COVID-19 pandemic, in contrast to face-to-face communication.53 Young people need genuine human interactions, reflecting God’s design of humans as relational beings (cf. Genesis 2:18). The more churches communicate with youth via AI-generated language, the less truly heartfelt those interactions become.54

As for the fourth argument, co-opting AI for sermon generation vastly differs from previous ministerial uses of technologies including the printing press. The printing press empowered humans to rapidly generate copies of human-authored words, requiring human study, ideation, and articulation. AI, however, allows humans to entirely bypass these essential processes of research, study, and articulation. Saying no difference exists between AI and the printing press is like saying no difference exists between using a microphone and hiring a speech writer. The former amplifies a person’s own words; the latter replaces them. (And in the case of AI, the replacement voice is not even human.)

Likewise for the fifth argument, an ethically significant difference exists between consulting books and commentaries and using AI for sermon generation (and not just sermon research). While consulting books, pastors engage their minds to study, glean insights, evaluate arguments, and draw connections between ideas. Pastors then link these ideas to key points distilled from Scripture for specific audiences, resulting in human-authored sermons and mentally sharpened shepherds. AI allows for bypassing all these steps. This is precisely why using AI for content-producing tasks saves time. AI cuts corners that, as the following section will argue, ought not to be cut.

In the end, none of these popular arguments withstands scrutiny. These lines of thinking cannot justify pastoral usages of AI that do not appropriately reflect Scripture’s portrayal of the responsibilities, habits, hallmarks, and heart of a shepherd.

Part Four: Arguments Against Using AI for Key Pastoral Tasks

Now we can turn to arguments against using AI for tasks that reflect the shepherd’s core responsibilities such as feeding, protecting, correcting, and tending the sheep.

Arguments Against Using AI for Personal Scripture Study

Although AI can analyze, summarize, and explicate Scripture, pastors should guard against the temptation to outsource their personal Scripture study to AI, for multiple reasons. The primary reasons are to stay close to God and to obey Scripture itself. Neglecting personal Scripture study contradicts biblical mandates for pastors, as seen in passages such as 2 Timothy 2:15 that emphasize diligence and personal responsibility in handling God’s Word. Relatedly, forgoing the study of Scripture sets a poor example for the flock, contradicting 1 Timothy 4:12 and 1 Peter 5:3.55

Outsourcing Scripture study to AI also undermines the cultivation of personal familiarity with God’s Word.

Outsourcing Scripture study to AI also undermines the cultivation of personal familiarity with God’s Word. Such familiarity is essential for daily pastoral life and ministry,56 as well as for cultivating a readiness to minister “in season and out of season” (2 Timothy 4:2).57^, 58 As Christensen observes, pastors also need thorough, ongoing knowledge of God’s Word to “lead learners to think biblically.”59

Additionally, refraining from outsourcing personal Scripture study to AI will help pastors promote biblical literacy. Christensen notes that biblical illiteracy is rampant in churches.60 If pastors do not diligently study Scripture due to relying on AI, their own biblical literacy will atrophy, leaving pastors less able to foster biblical literacy in their congregants.

Relatedly, maintaining personal Scripture study is vital for keeping pastors spiritually strong. Ministry is a spiritual war zone, and the only offensive weapon in the armor of God is the sword of the Spirit (Ephesians 6:17). Pastors who neglect personal Scripture study by outsourcing this task to AI loosen their grasp on this indispensable weapon. Therefore, coming to rely on AI for Scripture analyses, information, and insights places pastors in a spiritually dangerous situation.

Still another vital reason for pastors not to outsource personal study to AI is to maintain their intellectual acuity. Reliance on AI undermines pastors’ ability to think for themselves. Supporting this premise, a recent study from the Massachusetts Institution of Technology reveals how using ChatGPT to write essays negatively impacts the brain.61 Electroencephalography (EEG) of participants who leveraged ChatGPT exhibited “significant differences in brain connectivity” compared to participants who did not use AI.62 Namely, ChatGPT users showed “the weakest connectivity” and “struggled to accurately quote their own work,” leading the researchers to conclude, “While [large learning models] offer immediate convenience, our findings highlight potential cognitive costs.”63 Conversely, personal study (including of extrabiblical research sources) serves to exercise one’s mind, which is vital for the teaching, counseling, and problem-solving aspects of ministry.64

At this point, some may counter that using AI does not prevent pastors from studying Scripture; instead, AI can merely supplement personal Scripture study.65 However, a stated goal of sermon-generation software such as SermonDone is to save pastors from having to invest time in activities including personal study.66 The extent to which AI replaces studying God’s Word is the extent to which pastors lose the positives of personal Scripture study while simultaneously gaining the negatives of atrophied thinking skills. Outsourcing the study of sermon passages to AI also runs into other problems associated with using AI for sermon preparation.

Arguments Against Using AI for Sermon Preparation

While not an insurmountable problem, one potential issue cited in discussions about using AI in ministry is that AI does not always produce reliable content.67 AI-generated sermons have potential to feed the sheep with erroneous, biased, fabricated, or theologically problematic data.68 Critical review of AI-generated content can mitigate these concerns; however, studies on “automation bias” suggest that even professionals can tend to trust AI uncritically.69 Still, these issues are relatively minor compared to at least seven other reasons for choosing not to preach AI-generated sermon content.

First, Christians should lead the way in modeling intellectual integrity. Technically, using AI without citation is not plagiarism according to US copyright law.70 However, presenting AI-generated wording as though it is one’s own does not reflect high standards of honesty.71 Accordingly, academic institutions can consider this practice cheating.72 Yet honesty is necessary for one’s own character and conscience before the chief Shepherd.73 By practicing intellectual honesty, pastors can follow Paul’s exhortation for church leaders to “hold the mystery of the faith with a clear conscience” (1 Timothy 3:9).

Honesty and transparency are indispensable for maintaining trust with the congregation.

Honesty and transparency are indispensable for maintaining trust with the congregation. Ian Hussey observes, “In a similar way that congregations may be concerned about the preacher using Internet sermons or plagiarizing content from other sources, the use of AI to generate part, or all, of a sermon may be experienced as a breach of trust and a character failing.”74 (An intuitive test is to ask, “Would I be comfortable with the audience knowing this was AI generated?”)75

In an attempt to counter the concern that AI-generated sermons will upset congregants, the SermonDone website asserts, “Most people in your congregation are not interested in how you prepare your sermon as long as it is biblically sound, Spirit-led, and connects with their hearts.”76 Leaving aside the question of whether AI-generated sermons are Spirit-led, the only way to know if the congregation cares is to ask them. If SermonDone is correct that most people “are not interested,” then a pastor has nothing to lose by checking. If the pastor is up front about wanting to use AI, if the congregation accepts this usage, and if the church provides disclaimers to make new attendees aware of AI usages, then breaches of honesty, trust, and intellectual integrity can be mitigated.

Even so, imagine walking into a church where the pastor transparently announces, “Just so you know, about 40% of today’s message is AI-generated. Getting AI to structure my sermons and to provide points, illustrations, and exegetical insights gives me more time to engage with you in other ways throughout the week. Now, let’s open God’s Word and see what the algorithm had to say.” How would you respond? Likely, knowing this information up front would be preferable to thinking the sermon was entirely human-authored and then accidentally discovering it wasn’t. Still, wouldn’t it be more reassuring trust that the entire sermon, even if imperfectly phrased, came from a human heart? (In fact, the pastor who transparently presented the AI-generated church service later reported that the main congregational feedback was, “We miss your sharing from the heart.”77) While disagreements on this issue may persist, the remaining reasons listed below argue against even transparent uses of AI-generated sermons.

Turning to the second reason, choosing not to preach AI-generated materials will help pastors guard against laziness. The potential to become lazy through using AI is a temptation that pastors acknowledge.78 Hussey argues this potential poses a significant threat to the pastoral vocation.79 As Part One described, Scripture repeatedly portrays preaching sound doctrine as a labor that requires personal diligence, effort, and intentionality. The biblical virtue of diligence is a treasure too valuable to let AI corrode.

Third, opting not to preach AI-based messages helps pastors to “set the believers an example in speech,” as Paul commanded Timothy (1 Timothy 4:12). One commentator remarks that Timothy was to lead by the very “content, tone, and manner” of his communication.80 Using computer-generated speech undermines the pastor’s biblically mandated ability to lead by personal example in this way.

Fourth, choosing not to rely on AI for sermon preparation reflects the truth that a sermon’s purpose is to communicate God’s Word to this specific congregation. As Christensen aptly says, “God has a purpose for every unit of thought he has revealed in the Bible. Our job is to figure out what that purpose is for our people in this place and at this time, then deliver it to them.”81 Hussey notes, “Certainly, one could prompt generative AI to suggest a passage for a particular congregation at a particular time, but most preachers would recognize that the selection of a text is something that emerges from a deep pastoral knowledge of the congregation and prayerfulness, a task AI is incapable of performing.”82 To do this, the pastor draws on guidance from the Holy Spirit and responds to awe of God’s majesty, the comfort of his love, and the lavish riches of his grace.

Fifth, to resist the temptation to preach AI-generated sermons is to resist dehumanization in preaching and discipleship. Hussey warns that “perhaps the biggest ethical issue in the use of AI in preaching is the loss of the voice of the preacher in the sermon.”83 He aptly observes, “Only a human preacher can culturally exegete and empathize with their particular congregation and share the personal impact the message has had on their own being.”84 Another research team, despite generally advocating for AI, recognizes that AI “lacks the genuine human connection central to spiritual guidance and mentorship.”85

Sixth, preaching one’s own diligently prepared sermon reflects the value of Scripture, of the church, and of ministry itself. Through the sermon, God’s chosen shepherd unfolds God’s Words for God’s people, a high assignment that calls for appropriately high standards. Programs that encourage pastors to copy, paste, and preach AI-generated content suggest pastors do not need to invest serious thought into this assignment. By implying that unfolding God’s Word to God’s people is not worth serious consideration, using AI-generated sermons undermines a minister’s witness to the world, to the church, and to the minister’s own heart.

While chatbots can (at best) speak the truth, only a human pastor can speak the truth in love.

Seventh, preaching human-generated content matters because, while chatbots can (at best) speak the truth, only a human pastor can speak the truth in love (Ephesians 4:15). Love is the Christian’s most important task (Matthew 22:36–40), allowing our ministries to bear fruit (John 15:7–17). Preaching with efficiency, knowledge, and eloquence in the absence of love profits nothing (1 Corinthians 13:1–3). As Vinet says, “Preaching is a work of love, a good work, a good office, a part of the service of God.”86 Yet chatbots cannot love. AI can string together heartless words with breathtaking efficiency. But an efficiently clanging cymbal is still a clanging cymbal (1 Corinthians 13:1). To ensure that love permeates every part of the preaching process, pastors should not use AI to generate sermon content.

As a caveat, however, the foregoing discussion is not meant to imply that pastors cannot transparently incorporate AI outputs into sermons for illustrative or educational purposes. For example, a pastor might ethically say, “Here’s what ChatGPT answered when I asked it XYZ question. Notice how the chatbot said X, whereas Scripture says Y.” In this example, the pastor has (1) clearly labeled the AI’s contribution and (2) upheld his human responsibility as the congregation’s spiritual leader under the authority of God’s Word. The pastor has not outsourced any spiritual leadership responsibilities to AI.

Arguments Against Using AI for Pastoral Care

Along with Scripture study and sermon preparation, pastoral care represents another key set of tasks that pastors may be tempted to outsource. According to an article in Theology and Science, “Pastoral care refers to the help that clergy and trained laity provide to persons who are suffering, troubled, or perplexed, and specifically help grounded in a theological or spiritual perspective.”87 The article lists examples of existing AI systems designed for counseling contexts, including a Stanford psychologist-developed “talk therapy chatbot” known as the “Woebot.”88 Hypothetically, a version of the Woebot trained on biblical counseling data could “listen” to a troubled person’s concerns and provide advice similar to that which a human pastor may offer. But does the fact that pastors can outsource such care to chatbots suggest that they should?

In answer, Part One discussed how the pastor is the embodied, human shepherd of God’s embodied, human flock. Only a human can put the “pastor” in pastoral care. For that matter, only a human can put the “care” in pastoral care. Machines, by definition, cannot care. As one research team observes, “AI lacks true empathy, spiritual intuition, and the capacity for moral reasoning. Its responses, no matter how sophisticated, are generated from algorithmic processing and cannot embody the human essence of care, compassion, or spiritual insight.”89 To pretend that machines can care is to depersonalize the human relationships central to life and ministry and to misrepresent the profoundly personal, relational nature of God’s care for us. For all these reasons, pastors cannot appropriately forgo face-to-face pastoral care tasks by outsourcing these tasks to AI.90

Conclusion

In summary, a biblical theological evaluation of the church, the pastor, the sermon, and technology suggests that pastors cannot appropriately outsource key pastoral tasks to AI. Such tasks involve distinctly human elements such as love, care, worshipfulness, prayerfulness, and personal relationships with God and other humans, all of which go beyond AI’s capacities. Additionally, God’s Word calls human pastors to exercise personal responsibility, diligence, and integrity in shepherding God’s new covenantal community, a high calling that demands high standards. For all these reasons, AI cannot ethically replace pastors’ spiritual leadership responsibilities including personal Scripture study, sermon preparation, and pastoral care.

Pastors, however, can ethically use AI for purposes including certain research and non-content-producing administrative tasks, provided that high standards of intellectual integrity and transparency are maintained. Pastors can then face a dawning new age of church history not as purveyors of automated messages but as human shepherds of God’s embodied flock.

Artificial intelligence (AI) is rapidly transforming industries, and the pharmaceutical sector is poised to be one of its most significant beneficiaries, reshaping how drugs are discovered, tested, and brought to market. In a recent Bloomberg Television interview, Demis Hassabis, CEO of DeepMind and Nobel laureate, revealed that AI could dramatically reduce drug discovery timelines, potentially cutting years of labor-intensive research down to mere months. DeepMind’s advanced AI models aim to streamline the identification of promising drug candidates, enhance precision, optimize molecular design, and reduce the high failure rates that have historically plagued pharmaceutical development. This breakthrough promises faster access to innovative treatments, lower development costs, improved patient outcomes, and a transformative new era of medical research powered by sophisticated computational intelligence and predictive modeling.

How AI is changing the drug discovery process: DeepMind CEO reveals

Traditional drug discovery involves painstaking laboratory experiments, lengthy clinical trials, and significant trial-and-error testing, often taking 10–15 years from concept to market. According to Hassabis, AI can radically alter this timeline.“In the next couple of years, I’d like to see that cut down in a matter of months, instead of years,” Demis Hassabis said in an interview with Bloomberg Television. “That’s what I think is possible. Perhaps even faster.”DeepMind’s subsidiary, Isomorphic Labs, leverages AI to model complex biological systems, analyse molecular structures, and predict interactions between drugs and proteins. In the Bloomberg interview, Hassabis highlighted that AI can process enormous datasets far faster than human researchers, enabling the identification of promising drug candidates within weeks instead of years.This accelerated approach could not only save valuable time but also optimize resource allocation, ensuring that researchers focus on molecules with the highest likelihood of success.

How AI predictive models are transforming drug discovery and minimising setbacks

A major challenge in drug discovery is the high failure rate: many compounds that look promising in early tests fail in later stages due to inefficacy or harmful side effects. Hassabis emphasized that AI’s predictive capabilities could reduce these failures significantly.DeepMind’s models simulate protein folding and chemical interactions, allowing scientists to forecast how molecules behave in the body. The AI can also suggest novel molecular structures that traditional methods might overlook, expanding the pool of potential therapeutics. By prioritizing candidates most likely to succeed, AI improves efficiency and reduces costly setbacks in research.

AI’s role in speeding up drug development and expanding access

Hassabis discussed the broader implications of AI-driven drug discovery in the Bloomberg interview. Faster development cycles could allow for quicker responses to pandemics, emerging diseases, and critical health crises. Moreover, AI could facilitate the creation of personalized medicine, tailoring treatments to individual genetic profiles, metabolic rates, and disease characteristics.Beyond speed, AI’s efficiency could lower drug development costs, making treatments more accessible globally. This democratization of medicine could have profound social impacts, particularly for developing nations where access to cutting-edge therapies is limited.

From Alzheimer’s to rare cancers: AI leads the way

While Hassabis did not provide specific drug names in the interview, he emphasized that AI models are already being applied to several disease areas, including neurodegenerative disorders, rare genetic conditions, and chronic illnesses. Early studies suggest that computational predictions could significantly reduce the experimental burden and provide actionable leads for human trials.For instance, modeling protein-drug interactions can identify compounds that might mitigate protein misfolding in diseases such as Alzheimer’s. Similarly, AI-driven analysis of molecular pathways could accelerate treatments for rare cancers where conventional drug development is often economically unviable.

AI-driven drug discovery: Challenges

Despite its promise, AI-driven drug discovery is not without challenges. Hassabis pointed out several critical considerations:

Addressing these challenges will be crucial to translating AI’s predictive power into real-world therapies.Also Read | Abidur Chowdhury: Meet the designer behind Apple’s ultra-slim iPhone Air and its futuristic technology