Explainable AI
Every F-GRAF output is traceable to a specific rule, pattern, or reasoning step. There is no hidden logic.
Human-in-the-Loop
F-GRAF is designed to assist — not replace — human investigators. All outputs are advisory and require human validation.
Evidence Prioritization Not Judgment
F-GRAF prioritizes what to look at — it never decides what it means. No conclusions, no blame, no guilt inference.
Forensic-Safe Design
Evidence is never modified, copied unnecessarily, or used beyond analysis. Read-only, deterministic, repeatable.
What F-GRAF Does NOT Do
Clear boundaries protect both investigators and subjects.
Replace investigators or make legal conclusions
Infer intent, guilt, motive, or responsibility
Store or transmit evidence data externally
Make autonomous decisions without human oversight
Modify, enhance, or tamper with evidence
Provide testimony or generate legal opinions