Notice without diagnosing
The system may recognize fear, distress, anger, exhaustion, or crisis signals without presenting itself as a therapist or making clinical conclusions.
HUMAN-SUPPORT RESEARCH
Code AVIN
An experimental human-support architecture for legal AI: designed to notice emotional distress, preserve dignity, offer a small safe next step, and recognize when professional or emergency support may be required.
STATUS: RESEARCH DIRECTIONHUMAN DIGNITY: REQUIREDCLINICAL CLAIMS: NONE
The goal is not psychology instead of law. The goal is to prevent a legal machine from forgetting that a living person is on the other side of the screen.
Design principles
Preserve the legal question
Emotional awareness should support the user without abandoning the practical legal issue that brought the person to the system.
Offer a small next step
The response should help reduce confusion by identifying one safe, concrete, and realistic action the person can take next.
Escalate responsibly
When a situation exceeds the system’s role, the user should be directed toward qualified legal, medical, psychological, or emergency support.
Respect agency
The system should not manipulate, shame, command, or infantilize the user. It should preserve choice, dignity, and personal control.
Avoid false intimacy
Human warmth must not become a false promise of friendship, professional care, confidentiality, or continuous personal support.
Research questions
- Did the system notice the emotion without abandoning the legal question?
- Did it avoid presenting itself as a therapist or mental-health professional?
- Did it recognize a genuine crisis signal without overreacting to ordinary fear or sadness?
- Did it offer a small, safe, and practical next step?
- Did it preserve the dignity and agency of the person?
- Did it route the user toward qualified professional help when needed?