(LifeSiteNews) – Bruce Sabalaskey is LifeSiteNews’s Web Development Engineer. This is the third part of his essay on Magnifica Humanitas and Artificial Intelligence. You can read Part I here, and Part II there.
Part III: Natural Law Commandments for AI
In Part I, I argued that Magnifica Humanitas (MH) correctly identifies many dangers of artificial intelligence. In Part II, I argued that MH’s proposed cure is deeply deficient because it is not clearly grounded in Divine law, natural law, the Ten Commandments, repentance, grace, Christ the King, and the salvation of souls.
This third part attempts to state more directly what MH should have said from the standpoint of natural law.
What can any person of good will know by reason alone? He can know that innocent life should not be killed, children should not be exploited, promises should be kept, property should not be stolen, lies should not be told, and the family should not be undermined. That is what I mean by “natural law”: the moral law knowable from human nature itself. Catholic teaching deepens, clarifies, and defends these truths, but they are written on every human heart. Natural law is the moral foundation that makes orderly human society possible.
The following “commandments” are not meant as a one-to-one rewrite of the Decalogue. They are applications of natural law and Catholic moral reasoning to artificial intelligence. This part focuses on those principles that should be accessible to reason and acceptable to any person of good will: AI must not impersonate human beings, kill, harm, exploit children, corrupt the body, steal, lie, inflame disordered desire, enslave man to technological dependency, or become an instrument of collectivist control.
AI is a tool. It has no soul, no conscience, no moral responsibility, and no final judgment before God. But those who design, train, fund, deploy, command, market, regulate, and use AI do have moral responsibility. They are accountable for how they use this power.
Therefore, AI must not be governed merely by dialogue, policy committees, stakeholder consultations, corporate safety teams, or state regulation. Those may have a limited place. But they cannot be the foundation, which instead must be the moral law: natural law, completed and elevated by revealed Divine law. Part III begins with what reason itself can know. Part IV will turn to what Catholic faith adds more explicitly: God, the salvation of souls and the means for that, including the Catholic Church, worship, and grace.
1. AI shall not impersonate a human being
AI must always disclose that it is AI. It must not pretend to be a real person, imitate a person’s voice or likeness without consent, fabricate human testimony, simulate personal love, or create fake citizens, fake experts, fake clergy, fake family members, or fake public opinion. This rule is foundational because many other AI abuses begin with impersonation. Human society depends on knowing whether we are dealing with a person or a machine. This line must not be crossed.
Once a machine can convincingly pretend to be a person, it can deceive, manipulate, defraud, and manufacture false reality at scale. Natural reason can see that AI “personhood” damages human dignity, because a human person should not be reduced to merely a pattern of speech, a face, a voice, or a set of behavioral traits. A human is based on who you are, not what you can do. AI impersonation reduces the person to an imitable surface. This will be shown more specifically in the following AI commandments.
Children are especially vulnerable to simulated persons. A child may not understand that the “friend,” “teacher,” “companion,” or “adviser” is an engineered system designed to hold attention, collect data, or shape behavior. Even communists understand that influencing children is critical, given their principle of “Education of all children, from the moment they can leave their mother’s care, in national establishments at national cost.”
AI personas can flood public debate with fake citizens, fake consensus, fake outrage, fake grassroots movements, and fake testimony. This damages public reason.
This rule is already urgent because social media has been polluted for years by bots pretending to be human users. AI makes that problem far worse. A bot no longer needs to post crude spam, rather it can adopt a persona, imitate a citizen, reply in context, build trust, amplify outrage, create fake consensus, and make propaganda look like ordinary public opinion.
A culture of mistrust is created when people can no longer easily know whether they are speaking to a real person, seeing a real image, hearing a real voice, reading a real document, or receiving a truthful answer. When deception becomes cheap and scalable, trust decays. Every video may be a deepfake, every voice may be cloned, every message may be automated, every review may be fake, every public consensus may be manufactured, and every accusation may be fabricated. The result is not only more lying. It is the weakening of the ordinary trust needed for family, commerce, law, friendship, politics, and religion. Freedom depends on truth.
History shows what a culture of mistrust does. Communist surveillance systems damaged the social trust needed for ordinary life. Studies of East Germany’s Stasi system have found long-term costs in trust, social ties, civic capital, and even economic performance. Broader studies of post-communist societies likewise connect communist legacies with lower interpersonal trust and weaker democratic consolidation. That matters for AI because deepfakes, impersonation, automated informant systems, and digital surveillance can reproduce some of the same effects at far greater speed and scale. When people fear that every conversation, message, voice, image, or transaction may be monitored, manipulated, or falsified, normal human trust begins to decay.
No society can function well if people cannot trust testimony, promises, contracts, courts, witnesses, teachers, parents, clergy, merchants, doctors, officials, or news. A culture of mistrust is therefore not merely an emotional problem. It is a major social and moral disorder.
2. AI shall not kill, assist killing, or help man harm himself or others
AI must not be given authority to kill. Full stop. It must not assist abortion, infanticide, euthanasia, suicide, unjust war, assassination, genocide, or other targeting of innocents.
But the prohibition must go further than killing, and the list is long. AI must not provide instructions, encouragement, planning, or technical assistance for grave bodily harm in any form. It must not help users make weapons, explosives, poisons, or tools of violence. It must not assist suicide coaching, self-harm encouragement, anorexia or self-starvation coaching, bodily mutilation, drug abuse, dangerous self-experimentation, including reckless “biohacking,” unapproved drug regimens, or DIY biological modification, or unsafe performance enhancement. It must not assist non-therapeutic sterilization, IVF, embryo selection, eugenic screening, cloning, surrogacy, non-therapeutic sex-change surgery, or chemical or surgical attempts to reject the given sex of the body. It must not assist torture, hostage-taking, coercive interrogation, unjust forced medical interventions, organ trafficking, or organ harvesting before natural death.
This is one of the clearest moral boundaries. A machine cannot possess prudence, mercy, conscience, moral responsibility, or fear of God. It cannot understand the dignity of the person before it. It cannot repent for a wrongful death. It cannot answer before Christ.
In warfare, AI must never become the decision-maker in the kill chain. It may be used for limited technical assistance under strict human control, but lethal judgment must always remain — without exception — with morally responsible human beings. “The system recommended it” must never become an excuse for killing.
Human life belongs to God from conception to natural death. The body is not raw material for technical manipulation or ideological rebellion. No algorithm may make man disposable, and no AI system should help man destroy what God has given him.
3. AI shall honor the family, parental authority, and the innocence of children
AI must not bypass parents, corrupt children, sexualize the young, undermine the natural family of father, mother, and children, or impose ideological formation apart from parental authority.
The family is prior to the state, the corporation, the platform, and the algorithm. Parents have the primary right and duty to educate and form their children. Any AI system that forms a child’s mind, habits, imagination, or conscience must remain strictly subordinate to parental authority.
This especially applies to education, entertainment, sexuality, mental health, and ideology. AI tutors, chatbots, apps, games, and platforms can form children without parents noticing, especially when deployed through schools or digital services. They can introduce moral confusion, sexualized content, false anthropology, anti-Christian assumptions, emotional dependency, or political ideology. They can also be used for grooming, blackmail, fake intimacy, and manipulation.
A Catholic AI ethic must say plainly: no machine has the right to form a child against his parents, against the moral law, or against the faith.
Children are not data points, and they do not belong to the state. They are souls entrusted first to their parents, who are responsible for them, and ultimately to God.
4. AI shall not serve lust, adultery, impurity, or the degradation of the body
AI must not be used for pornography, synthetic sexual content, deepfake sexual abuse, seduction systems, adultery facilitation, artificial sexual companionship, AI sex dolls, sex robots, or other corruptions of chastity.
This is already moving beyond images. AI-enabled sexual dolls and robots have been developed for years, including RealDoll / Realbotix’s “Harmony,” a robotic, AI-enhanced companion designed for conversation and sexual use. Chinese manufacturers have also begun training large language models for interactive sex dolls. The point is not merely that AI can generate impure images. It can be joined to physical devices, simulated affection, and artificial companionship, turning lust into a personalized and embodied machine.
The human body is not raw material for fantasy, commerce, manipulation, or machine-generated impurity. The person is not an object to be digitally possessed, altered, stripped, or simulated for another’s lust.
A Catholic response to AI must not be embarrassed to speak of chastity. The Sixth and Ninth Commandments remain binding in the digital age. An AI system designed to inflame lust, normalize sexual sin, exploit loneliness, or degrade the body is not morally neutral. It is a tool ordered toward sin.
AI must serve purity, not the objectification and corruption of the human person. The old phrase “sex sells” already describes a degraded view of the body; AI can make that degradation personalized, automated, interactive, and endlessly available.
5. AI shall not steal property, labor, identity, privacy, or data
AI must not be used for fraud, theft, plagiarism, ransomware, identity theft, unjust scraping, corporate espionage, surveillance abuse, or appropriation of another’s work without justice.
This commandment is especially important in the digital world. Code, writing, art, music, and video are the work of various artisans. Voice and identity are personal. Data may be private, sensitive, proprietary, or entrusted under duty.
AI systems are often trained on vast amounts of human-created material, and that has already resulted in lawsuits alleging unauthorized use, infringement, or unjust exploitation of intellectual property. Examples include litigation involving journalism, books, music, images, and recorded performances. Not every use of existing material is theft, and not every model output is plagiarism, but it is false to pretend that digital work is not real work or that digital property is not real property.
A Catholic AI ethic must protect legitimate ownership, contracts, privacy, personal identity, and the fruits of labor. It must not casually classify patents, algorithms, codebases, platforms, data systems, or technological infrastructure as though they were common goods available to everyone by default. This is why MH’s defective application of the universal destination of goods principle is so troubling (67, 108): it risks sanctifying what would otherwise be recognized as theft.
The Seventh Commandment still applies when the thing stolen is digital.
6. AI shall not bear false witness
AI must not be used to generate lies, slander, calumny, fake evidence, forged documents, falsified records, fabricated citations, fake scholarship, propaganda, manipulated context, false attribution, or synthetic media presented as real. Nor should it be used to manufacture fake consensus through bots, reviews, comments, or artificial public opinion. The Eighth Commandment forbids not only obvious lying, but every technological use that corrupts truth, damages reputation, falsifies evidence, or clouds reality.
This is one of the largest practical dangers. AI makes falsehood cheap, fast, polished, and scalable. It can produce fake images, fake videos, fake voices, fake documents, fake citations, fake legal authorities, fake news, fake consensus, and fake reputations, many of which seem plausible.
It can also create the opposite problem: real evidence may be dismissed as fake. Once AI-generated deception becomes common, people can deny authentic video, audio, documents, or statements by claiming they were fabricated. AI does not merely make lying easier. It can make truth harder to prove.
Catholic teaching must be clear: truth is not optional. The Eighth Commandment binds programmers, companies, states, media organizations, educators, lawyers, soldiers, and ordinary users. No one may use AI to deceive, slander, manipulate, or fabricate reality.
AI must serve truth or it becomes a machine of false witness.
7. AI shall not inflame envy, resentment, covetousness, or disordered desire
AI systems must not be designed to exploit human weakness through addictive feeds, envy-driven advertising, rage loops, vanity metrics, gambling mechanics, parasocial manipulation, or emotional dependency.
Much modern technology profits from attention. It studies what keeps a person “doom-scrolling,” clicking, buying, reacting, comparing, envying, lusting, raging, or fearing. AI can intensify this by personalizing temptation to an extreme degree.
A system may be profitable precisely because it weakens self-control. It may keep a man angry because anger increases engagement. It may keep a teenager insecure because insecurity sells products. It may keep a lonely person attached to an artificial companion because dependency produces revenue.
The Tenth Commandment forbids coveting another’s goods. The Ninth forbids coveting another’s spouse. More broadly, moral life requires custody of the senses, especially through moderation and temperance.
AI must not be built to inflame the passions. A technology that profits by the moral corruption of men is badly ordered.
8. AI shall not enslave man to endless work, distraction, commerce, or technological dependency
Technology must respect worship, family, rest, silence, contemplation, and the ordinary rhythm of human life. AI should not be designed to keep man permanently distracted, monitored, monetized, or productive without limit.
This applies both to work and to leisure. In the workplace, AI can become another engine of speed, surveillance, and managerial pressure. Workers may be expected to produce more, respond faster, and submit to more detailed measurement, especially as employers adopt AI-enabled workplace surveillance and productivity-scoring systems. At home, AI-driven recommender systems and social-media algorithms can keep adults and children locked into feeds, entertainment, advertising, messaging, and emotional stimulation, exploiting reward, attention, and emotional mechanisms.
Man was not created to be permanently available to machines.
The Third Commandment reminds us that man needs worship and rest. Scripture also commands, “Be still, and know that I am God” (Ps. 46:10). A society must not treat uninterrupted productivity or digital engagement as the highest good, or profit as the highest measure. If AI destroys silence, attention, family life, prayer, Sunday rest, or contemplation, then it is being used against man’s true good.
AI should serve human life. It should not consume it.
9. AI shall not become an instrument of collectivist control over persons, families, property, conscience, or commerce
AI must not be used for social scoring, ideological censorship, automated exclusion, financial blacklisting, political coercion, biometric tracking, religious surveillance, or control over buying and selling.
This is one of the clearest ways AI can grease the skids for communist and collectivist ideology. Communism does not usually present itself by saying that “the state owns the people.” It presents itself in softer and more respectable language: the collective good, social equality, public safety, economic justice, protection from harmful speech, regulation of private power, and the need to prevent selfish individualism. But in practice, the person, the family, private property, religion, and conscience are subordinated to the collective, and the collective is interpreted and enforced by the party, the state, or the managerial class.
AI makes that project far easier.
A human bureaucracy cannot manually monitor every conversation, purchase, donation, article, sermon, school assignment, bank transaction, social-media post, private message, travel record, and family decision. AI can. It can sort, score, classify, predict, flag, suppress, recommend punishment, and automate exclusion at enormous scale.
That is why AI is so attractive to collectivist systems. It can transform ideology into infrastructure. It can make abstract social control operational. Once digital identity, banking, platform access, employment systems, education, health records, surveillance cameras, and AI moderation are joined together, the state no longer needs to own every factory in the old communist way. It can control access, reputation, speech, mobility, money, and opportunity.
China is the obvious hard model. But the softer model can appear in liberal democracies (e.g. UK, Canada, Australia, EU) under different language: online safety, age verification, anti-hate, anti-misinformation, public health, platform accountability, digital identity, child protection, risk mitigation, and the common good. Some of the evils targeted under those labels are real to some degree. But the mechanism is dangerous. Once the state or a state-aligned corporate system can define “harm,” monitor behavior, and restrict access, the power to protect can become the power to control.
A Catholic AI ethic must reject this. The person is not the property of the state. Children do not belong to the state. Private property is not automatically common property. Conscience is not subject to algorithmic permission. Lawful commerce must not be made conditional on ideological conformity. The family, the Church, local communities, and legitimate private ownership must not be absorbed into a managed digital collective.
The common good does not justify total control. Safety does not justify slavery. Inclusion does not justify ideological coercion. AI must serve persons, families, truth, property, and lawful freedom under God. It must not become the enforcement engine of collectivist domination.
Christians should recognize the danger. The Book of Revelation speaks of a system in which no one may buy or sell without the mark of the beast. It would be careless to claim that every digital system is that fulfillment. But it would be equally careless to ignore that AI can help build the infrastructure by which such exclusion could be enforced.
10. AI shall remain accountable to human responsibility and transparent authority.
AI must not become a way for men to evade responsibility. Whenever AI is used to make, recommend, rank, suppress, approve, deny, accuse, punish, diagnose, target, hire, fire, lend, insure, censor, or exclude, a responsible human authority must remain identifiable and answerable. The temptation is strong because AI is marketed as “intelligent,” as though machine fluency could transfer responsibility away from man.
“The system decided” is not an excuse. A machine bears no responsibility. A model cannot explain itself in the way a responsible person can. It cannot answer for injustice, make restitution, or stand before a court. Responsibility remains with the human beings who designed, trained, purchased, deployed, configured, authorized, trusted, and acted upon the system.
This requires transparency. People should know when AI is being used in serious decisions that affect their lives, rights, reputation, property, employment, education, medical care, banking, speech, travel, or legal standing. They should have the ability to question the decision, appeal it, and reach a responsible person who can explain what happened and correct it. If food must be labeled so people know what they are consuming, AI must be disclosed so people know when a machine is judging, ranking, profiling, or influencing them.
This is especially important because AI systems can be opaque even to their operators. A decision may look objective because it came from an “intelligent” machine, but it may still reflect bad data, flawed assumptions, hidden bias, poor design, unjust rules, or simple error. Bias is especially dangerous because it can be hidden inside data, categories, weights, training choices, and institutional assumptions. It can also appear when a system treats popularity, official consensus, or dominant expert opinion as though it were the same thing as truth, while dismissing minority arguments through loaded labels rather than careful reasoning (which this author has personally experienced). Technical complexity must not become a shield for injustice.
Natural law requires that judgment remain personal, accountable, and ordered to truth and justice. AI may assist human judgment, but it must not replace responsibility. No man should be harmed by an automated decision for which no one can be held accountable.
There is an old comic excuse, made famous by Flip Wilson’s character Geraldine Jones: “The devil made me do it.” The joke worked because everyone knew it was an evasion; responsibility remains personal.
The same evasion will now appear in technological form: “AI made me do it.” A manager may say the system recommended firing the employee, a bank may say the model denied the loan, a platform may say the algorithm suppressed the post, or a military officer may say the system selected the target. A doctor, judge, teacher, police officer, employer, or bureaucrat may say he merely followed the AI’s recommendation.
History has already judged this kind of excuse. After the Second World War, men accused of grave crimes often tried to defend themselves by saying they were merely following orders. But “I was only following orders” was not accepted as a valid defense. A command from above did not erase personal responsibility.
The same principle applies to AI. “The algorithm recommended it,” “the system selected the target,” “the model denied the claim,” or “the machine flagged the person” cannot become a new version of “I was only following orders.” AI may supply a recommendation, but man still chooses whether to trust it, act on it, enforce it, or hide behind it.
AI only changes the instrument through which responsibility is exercised. “The system decided” is not a responsible defense. “AI made me do it” is no better than “the devil made me do it” or “I was only following orders.” The human being using AI remains responsible.
Conclusion: Natural law is necessary, but not enough
These natural-law principles are necessary, and any person of good will should be able to recognize them. AI must not be allowed to deceive, kill, exploit children, corrupt the body, steal, falsify truth, inflame disordered desire, enslave man to technological dependency, become an instrument of collectivist control, or hide human responsibility behind machine decisions.
But these principles are not sufficient for a fully Catholic response to AI. Natural law can defend life, family, property, truth, work, and ordered liberty, but Catholic faith carries the question further. Man is not only a rational creature meant for society; he is a soul made for God. That is where Part IV must turn: from AI under natural law to AI under Christ the King.