Start with the purpose

Digital radionics should be presented as experimental symbolic software, not as a medical diagnostic or treatment system. A clear purpose statement protects users from confusing personal reflection with professional healthcare or other regulated advice.

Consent before capability

Software may technically allow a user to enter another person’s name, image or details. That does not make every use appropriate. Personal data should be handled with permission, and symbolic work involving another person should respect their autonomy rather than attempt control or coercion.

Privacy is part of the design

Witness images, names, birth information and session notes can be sensitive. A responsible application should collect only what is needed, explain where data is stored, support deletion and avoid unnecessary cloud transfer.

Health claims require a hard boundary

No symbolic score, matrix or AI-generated interpretation should be presented as a diagnosis. Users with health concerns need established medical assessment. The same principle applies to legal, financial and psychological decisions where professional expertise may be required.

AI needs its own boundary

AI can improve wording or summarise notes, but fluent language can create an illusion of authority. Generated text should be identified as interpretation. The model should not invent measurements, hidden causes or medical conclusions that are not present in the data.

Vulnerable users require more caution

Applications should avoid reinforcing paranoia, compulsive checking or the idea that every coincidence is a meaningful external signal. Design can help by limiting certainty language, encouraging breaks and reminding users to consider ordinary explanations.

Audit trail and reversibility

If an application runs automated sessions, generates content with AI or uses schedules, the user should later be able to understand what happened. A useful audit trail records the time, selected mode, text used, algorithm version and important settings. It should also be possible to stop, delete or cancel a session without hidden consequences.

Data minimisation

Not every piece of data that can be collected is necessary. If a feature works without a name, date of birth or other personal information, there is no reason to request it. Data minimisation reduces the impact of a breach and makes it clearer which information actually serves a functional purpose.

Extra caution with AI-generated interpretation

A language model can produce a persuasive explanation even when the input data do not support such a strong conclusion. The interface should therefore distinguish measured data, symbolic interpretation and AI commentary. For sensitive topics, uncertainty and alternative explanations are preferable to a confident conclusion without evidence.

Red flags for responsible design

  • hiding content from the user or from another person,
  • promising health effects without appropriate evidence,
  • creating fear that sessions must continue,
  • encouraging dependence on the application for important life decisions,
  • unclear storage of personal data or no practical deletion path.

A responsible system does not turn these patterns into marketing advantages; it designs against them.

Not every feature carries the same risk

A visual symbol created by a user for personal reflection has a different risk profile from a feature that secretly presents messages to another person, stores sensitive personal data or generates health predictions. Features should therefore be assessed by their possible consequences, not only by technical novelty.

The higher the potential risk, the stronger the need for clear explanation, consent, reversibility and limits on automation.

Safe defaults matter

A responsible application should be safe in its default state. Experimental features should remain off until the user deliberately enables them. Intention text should remain visible, data sharing should be optional, and AI should not generate health or other sensitive conclusions without clear context and boundaries.

These defaults reduce accidental misuse while still allowing informed users to explore experimental features deliberately.

Ethical intention design

A useful intention focuses on the user’s own behaviour or desired state: clearer communication, disciplined preparation, calm attention. An ethically weak intention attempts to force another person to act, guarantees an outcome or frames harm as a goal.

Professionalism means showing uncertainty

The safest system is not one that promises the most. It clearly labels what was entered, what the software generated, what was measured and what remains a personal interpretation.