Public Technology, Human Dignity, and the Image of G-d
A government sees a citizen through categories.
Age. Address. Income. Tax status. Residency. Employment. Medical eligibility. Educational history. Legal status. Benefits received. Property owned. Licences held. Applications submitted.
These categories are necessary. Without information, a government cannot distribute benefits, collect lawful revenue, plan infrastructure, investigate wrongdoing, protect public safety, or understand the population it serves.
But a category is not a person.
A data record may describe certain facts about a human being. It cannot contain the human being’s complete reality. It does not record the soul, the hidden struggle, the moral potential, the private act of kindness, the prayer, the regret, the courage required to rise after failure, or the mission for which that person was created.
The Torah’s declaration that humanity is created b’tzelem Elokim—in the image of G-d—means that human worth is conferred by the Creator. It is not issued by a government, confirmed by an algorithm, or calculated from economic usefulness.
This does not mean institutions should abandon data.
It means data must remain beneath human dignity.
Why Governments Need Information
The responsible use of information can save lives.
Public-health systems need reliable records. Emergency agencies need to know where vulnerable populations live. Infrastructure authorities need demographic projections. Courts need accurate evidence. Tax systems require financial information. Benefits programs must determine eligibility. Security agencies may require lawful intelligence to confront genuine threats.
Refusing to collect necessary information can be as irresponsible as collecting too much.
A government that does not know which bridge is failing, which medicine is unavailable, which community lacks clean water, or which benefits are being stolen cannot serve effectively.
The moral question is therefore not whether a government may possess information.
It is:
Which information does it legitimately need, for what purpose, under whose authority, with what protections, for how long, and subject to what correction or appeal?
These questions transform data collection from an appetite into a discipline.
Without them, institutional curiosity expands naturally. A database created for one purpose is linked to another. Temporary emergency information becomes permanent. A convenient field is added because it may be useful someday. A commercial provider offers new inferences that no one originally requested.
The institution gradually moves from knowing what it needs to know toward knowing whatever it is technologically capable of knowing.
Capability, however, is not authority.
The Map Must Not Replace the Person
Data is a map.
A good map can help a government navigate reality. A distorted map can direct public power toward the wrong person, place, or problem. Even a highly accurate map remains a representation rather than the territory itself.
Institutional danger begins when officials forget the difference.
A person is classified as low risk, high risk, eligible, ineligible, compliant, delinquent, verified, suspicious, vulnerable, productive, inactive, or unlikely to succeed.
Some classifications may be administratively necessary. But every classification selects certain facts while excluding others. It compresses a multidimensional person into the variables relevant to a particular institutional decision.
The classification should therefore remain narrow.
A finding that someone is ineligible for one benefit does not mean the person is undeserving. A financial risk score is not a moral verdict. A prior mistake is not a complete identity. A government-domain email address does not establish official governmental endorsement. A citizen’s economic contribution does not determine the value of the citizen’s life.
The glossary concept of a “false verdict identity” describes a distorted self-concept mistaken for the whole person. Governments can commit a parallel error: an administrative verdict can be treated as though it has captured the person’s essence.
It has not.
Only G-d knows the whole person.
Classification Creates Power
To classify is to exercise power.
A category can open a door or close it. It can release a payment, trigger an investigation, deny an application, assign a priority, increase scrutiny, or shape how every later official interprets the person.
Even an apparently neutral label can acquire consequences far beyond its original purpose.
Suppose an institution labels an account as “government-domain activity.” That statement may be technically accurate if the activity is associated with a governmental email domain. But it does not prove that the government formally approved, purchased, sponsored, or endorsed the activity.
If the narrower fact is displayed as though it establishes the broader conclusion, information becomes geneivat daat (creating a false impression / theft of knowledge)—the creation of a false impression.
Transparent classification should therefore state what has actually been verified and what has not.
A public leaderboard may legitimately report attributable commercial or learning activity. But it should distinguish:
- verified transactions from estimates
- issued learning from completed learning
- personal activity from institutional activity
- domain classification from official authorization
- financial contribution from spiritual worth
- participation from endorsement.
The credibility of a ranking depends less upon how impressive it appears than upon how honestly its boundaries are described.
A Score Is Not Judgment
Scores are attractive because they simplify comparison.
They allow institutions to rank applications, countries, suppliers, programs, risks, and performance. Properly designed, a score can make complex information more understandable.
But a score contains a hidden temptation: because it is numerical, it appears final.
Every score rests upon a moral architecture.
Someone decided which activities count. Someone selected the weights. Someone chose the time period. Someone determined whether one hour of learning, one unit of commerce, one verified session, or one institutional purchase should contribute to the result.
The formula may be mathematically correct while the selected meaning remains open to examination.
A leaderboard intended to encourage constructive national participation can be valuable. It can make hidden generosity visible, create positive competition, and inspire institutions to increase their contribution to learning.
But it must never imply that G-d’s judgment has been converted into ranking units.
Heaven cannot be audited by a database.
A country with a high score may still possess grave injustices. A country with a low recorded score may contain extraordinary private righteousness that the system cannot see. The leaderboard can measure defined activity. It cannot measure the total spiritual standing of a nation.
Honesty about this limitation does not weaken the system. It protects the system from idolatry.
When Measurement Becomes Identity
A government needs performance metrics. Without measurement, programs can consume resources while accomplishing little.
Yet metrics become dangerous when people begin serving the measurement rather than the purpose.
A school can raise reported performance while neglecting students who are difficult to measure. A hospital can optimize waiting-time statistics by changing how cases are recorded. An agency can reduce documented complaints by making complaints harder to submit. A learning program can report issued time as though it were verified participation.
This is the institutional form of externality without inward truth.
The metric is polished. The purpose has been abandoned.
Chassidic thought distinguishes between the garments of the soul—thought, speech, and action—and the soul itself. Garments matter enormously because they are how the soul enters the world. But a garment must express the person rather than replace the person.
Institutional metrics are garments. They express selected aspects of performance. When officials worship the garment, the living purpose disappears beneath the report.
The question should always return:
What reality was this measurement created to serve, and is it still serving it?
Daat Is More Than Stored Information
Chochmah (wisdom / the flash of insight) perceives an essential insight. Binah (understanding) develops and analyzes it. Daat (integrated knowledge / binding awareness) binds the knowledge to the person so that the truth begins to govern emotion and action.
A government database may contain enormous quantities of information while the institution itself lacks Daat.
It may know that a population is suffering but remain emotionally and operationally disconnected from the suffering. It may possess repeated warnings about a failing system without allowing the warnings to alter policy. It may record the number of people denied a service without asking whether the denial process is just.
Information becomes Daat only when knowledge creates responsible connection.
The teachings in The Art of Education distinguish raw information from knowledge that is integrated and made active. Data must be accurate and thorough, but it must also serve realistic understanding and constructive action.
For government, this means that dashboards cannot substitute for encounter.
Senior officials should sometimes hear directly from the people represented by the dashboard. System designers should observe the process as citizens experience it. A denial letter should be read from the perspective of the family receiving it. An identity-verification process should be tested by those who lack perfect documentation, technological confidence, or physical access.
Daat does not abolish institutional distance. It prevents distance from becoming indifference.
Privacy Is Not Secrecy Without Limit
Privacy is sometimes described as the right to hide.
That description is incomplete.
Appropriate privacy protects the conditions in which dignity, trust, family life, thought, communication, and personal development can exist. The glossary treatment of tzniut includes appropriate privacy and dignified conduct. Its detailed Jewish laws are Halachic and should not be carelessly converted into universal data regulations, but the underlying recognition is important: not everything true must be exposed to everyone.
Privacy is also not absolute.
A person cannot invoke privacy to prevent every lawful investigation, conceal every financial obligation, or obstruct the protection of life. Governments may possess legitimate reasons to collect and share information under law.
The moral discipline lies in proportion.
Information should not be collected merely because it is interesting. It should not be shared merely because another department requests it. It should not be retained forever merely because storage is inexpensive.
The closer information reaches toward a person’s health, family, finances, communications, beliefs, movements, or vulnerabilities, the stronger the justification and safeguards should become.
The Limits of Consent
Institutions often rely upon consent.
Consent matters, but it must be interpreted honestly.
A citizen applying for an essential government service may have little practical ability to refuse the accompanying data terms. A person facing unemployment, illness, legal difficulty, or loss of benefits is not negotiating from the same position as a consumer choosing between entertainment services.
A checkbox can record agreement without establishing meaningful understanding.
This does not invalidate every consent-based process. It means government should not use formal consent to escape substantive responsibility.
A respectful notice should explain, in understandable language:
- what information is being collected
- why it is required
- whether providing it is mandatory
- how it will be used
- with whom it may be shared
- how long it will be retained
- what rights of access, correction, or appeal exist
- what happens if the person refuses.
Clarity is an expression of truth. Deliberate obscurity is not made moral by a signature.
Collect What Is Needed
Gevurah (restraint and boundary) is the power of boundary.
In data governance, Gevurah asks the institution to stop.
Do not collect the additional field without a defined purpose. Do not grant every employee broad access. Do not retain obsolete records merely because deletion requires effort. Do not combine databases before examining whether the new inference is lawful, fair, and necessary.
Restraint is not hostility to knowledge. It protects knowledge from becoming domination.
A disciplined public-data system should consider:
- purpose limitation
- data minimization
- role-based access
- authentication and security
- accuracy and correction
- retention schedules
- lawful deletion
- audit trails
- vendor restrictions
- incident response
- independent oversight
- meaningful appeal.
The exact legal requirements differ by jurisdiction. Torah does not provide a ready-made modern privacy statute. Qualified legal, technical, security, and Halachic professionals remain necessary.
Torah contributes the moral conviction that information is never detached from the dignity and responsibility of the people it concerns.
Accuracy Is a Form of Justice
Inaccurate public data can cause material harm.
A mistaken identity match can trigger an investigation. An outdated address can prevent a person from receiving notice. An incorrect financial record can block assistance. A defective family-status field can affect eligibility. A falsely elevated risk score can create repeated suspicion.
A system that collects data must therefore provide a realistic means of correction.
This is especially important when several agencies reuse the same record. One error can travel silently through multiple systems, accumulating the appearance of confirmation merely because the same original mistake appears in many places.
Repetition is not verification.
Justice requires more than accurate intake. It requires the humility to permit challenge.
Hod (humility / acknowledgment) is the faculty of acknowledgment. In institutional life, Hod allows the government to say: “The record was wrong. The model failed. The classification overstated what was known. The decision must be corrected.”
An appeal process is not an insult to administrative authority. It is one of the mechanisms through which authority remains worthy of trust.
Automation Must Remain Answerable
Automated systems can assist governments in processing applications, detecting anomalies, allocating inspections, identifying patterns, and improving the consistency of routine decisions.
They can also create a dangerous distance between power and responsibility.
An official may assume that a system is objective because it uses mathematics. A vendor may claim that the model is proprietary. A citizen may be unable to discover why an adverse result occurred. The government may continue using the system because replacing it would be costly.
Yet no automated system entered public service by itself.
Human beings selected it. Human beings authorized its purpose. Human beings accepted its limitations. Human beings remain responsible for supervising its effects.
The more consequential the decision, the more important it becomes to provide:
- validated data
- defined performance standards
- testing for foreseeable error
- monitoring after deployment
- human review with actual authority
- understandable reasons where required
- a route to challenge the result
- suspension procedures when harm emerges.
“Computer says no” is not justice.
The machine may calculate. It cannot bear moral responsibility.
Government Vendors Share the Data Obligation
Data responsibility does not end when information is transferred to a contractor.
A government cannot release sensitive records to a commercial provider and then treat the provider’s systems as morally external. The public did not entrust its information to an invisible abstraction called “the supply chain.”
Contracts involving data should clearly govern:
- permitted use
- security requirements
- subcontractor access
- cross-border handling where relevant
- model training or secondary commercial use
- breach notification
- record return or deletion
- audit rights
- incident cooperation
- termination obligations.
The government must also examine whether it will become unreasonably dependent upon a provider whose system it cannot understand, migrate, or supervise.
Technical sophistication should serve public stewardship. It should not create a new form of institutional helplessness.
Emergency Data Must Not Become Permanent Appetite
During an emergency, governments may need information they would not ordinarily collect.
A genuine danger to life can justify exceptional measures. But exceptional collection should be attached to a defined emergency purpose, restricted access, review, and an endpoint.
Otherwise, a temporary necessity becomes a permanent expansion of visibility.
When the emergency ends, officials should ask:
- Which data remains necessary?
- Which access permissions should be withdrawn?
- Which records should be deleted or archived under law?
- Which emergency integrations should be disconnected?
- Were any populations harmed disproportionately?
- What oversight was suspended, and when will it return?
Pikuach nefesh (the preservation of life) is a profound Jewish legal principle protecting life. It must not become a slogan used to eliminate every boundary. Claims of emergency require competent legal and Halachic judgment.
For non-Jewish governments, protection of life under the Noahide prohibition of murder remains fundamental, while the rules governing emergency data use arise through the country’s lawful system of justice.
The lanes remain distinct. The moral demand for truthful necessity applies to both.
The Ten Faculties of Public Data Stewardship
The ten faculties of the soul offer a framework for governing information.
Chochmah (wisdom / the flash of insight): What essential public purpose justifies the system?
Binah (understanding): Have its technical, legal, social, and human consequences been developed in detail?
Daat (integrated knowledge / binding awareness): Do officials remain connected to the people represented by the records?
Chesed (lovingkindness / expansive kindness): How will the information help, protect, educate, or enable service?
Gevurah (restraint and boundary): What must not be collected, exposed, retained, inferred, or automated?
Tiferet (harmonizing compassion): Does the system integrate public benefit with dignity and accuracy?
Netzach (endurance / perseverance): Can security, correction, and oversight be maintained throughout the system’s life?
Hod (humility / acknowledgment): Can the institution acknowledge error and accept legitimate challenge?
Yesod (bonding foundation): Is information shared only through trustworthy, defined channels?
Malchut (responsible implementation): Does the system produce just action rather than merely impressive analysis?
When Chochmah is absent, the system has no worthy purpose. When Gevurah is absent, collection expands without boundary. When Daat is absent, citizens become abstractions. When Hod is absent, errors become permanent. When Yesod is corrupted, information flows toward misuse. When Malchut is distorted, the data becomes an instrument of domination.
Technology does not replace the soul’s faculties. It magnifies whatever is governing them.
A Public Data Examination
Before launching or expanding a consequential government system, ask:
- What exact public purpose requires this information?
- What legal authority permits its collection and use?
- Is every requested field necessary?
- Could the system create a false impression about a person, institution, or nation?
- What important reality does the classification exclude?
- Can an affected person understand and challenge the result?
- How will errors be corrected across connected systems?
- Who can access the information?
- What may contractors and subcontractors do with it?
- When will the data be deleted, restricted, or reviewed?
- What human decision-maker remains accountable?
- Does the public presentation clearly distinguish measurement from moral worth?
These questions do not prevent innovation. They make innovation worthy of trust.
A Person Remains Greater Than the Record
A record can be corrected.
A score can change.
A classification can expire.
A human being remains a creation of G-d.
Sanhedrin 37b teaches the extraordinary significance of a single human life and the uniqueness of each person. Humanity begins with one person, yet no two people are entirely alike. This warns against reducing the individual to a replaceable unit.
Government must count because government must plan.
But it must count people without treating people as nothing more than what can be counted.
The database should help public servants see need more accurately. It should never train them to stop seeing the human being.