Hön na Kontenü
AGIRight.org

AILP — Protocole ti Yëngö ti Mändängö ti IA

Protocole ti Yëngö ti Mändängö ti IA

IA alîngbi tî mändä na ye so? Na yongoro wa, ndali ti sarango wa, na na ndïä wa?

Draft v0.1.1 Schéma ↓

01 · Nda-tënë

AILP operationalizes the AIRS spectrum for the specific question of learning. It distinguishes acts that binary crawler rules collapse together: being read is not being learned from; being retrieved is not being trained on; fine-tuning is not distillation; verbatim memorization is different from statistical internalization. AILP lets a publisher declare, in a single machine-readable file, per-dimension permissions — access, indexing, inference input, embedding, training, fine-tuning, distillation, memory, output, attribution and compensation — each as allowed, denied, or license-required. An AILP grant answers "may AI learn from this content?" only — it does not answer "may this Agent invoke a tool, write to a database, or publish?" (AARS), nor "who authorized this Agent?" (AADP).

02 · Ndâlï

  • Kîri tënë na hunda «IA alîngbi tî mändä na ye so?» na polêlê ti undêrê oko oko, aga na bit ti crawl gï pëpe.
  • Kângbi yëngö ti mändängö (lïngö na yâ ti modèle) na yëngö ti lêgë (gïngö apage).
  • Amodèle ti Futa so Andïä na Maboko: senge, aga ti mbongo pëpe, ahunda licence, kangbi-ngêrë.
  • Bâa so zïngö ayeke ye so masïni alîngbi tî sâra kua na ndö ni — ti si nzönî-bê agirisa aga bê-ûse ti ndïä pëpe.

03 · Portée

00

lêgë / index — gïngö na index ti nda-tënë

01

inference_input — sarango tongana ndo ti tënë na ngoi ti inference (RAG)

02

embedding — gbïängö aga vecteur na index ti wärängö

03

training / fine_tuning / distillation — anivo ti lïngö na yâ ti modèle

04

verbatim_memory — tongana a mû yëngö tî sâra kîri ni tâ tâ

05

fängö-iri / futa — kua ti citation na ahunda ti futa

04 · Tapande so masïni alîngbi tî diko

Wärängö ti yëngö ti mändängö — /ai/rights-spectrum.json (profil AILP)

/ai/rights-spectrum.json
{
  "version": "0.1",
  "protocol": "AILP",
  "publisher": "example.org",
  "default": {
    "access": "allowed",
    "indexing": "allowed",
    "inference_input": "allowed",
    "embedding": "allowed",
    "training": "license_required",
    "fine_tuning": "license_required",
    "distillation": "denied",
    "verbatim_memory": "denied",
    "attribution": "required",
    "compensation": "contact"
  },
  "contact": "licensing@example.org"
}

05 · Limites

  • Nda-tënë ti yëngö ti mändängö ayeke ngangü ahon ti bâa nzönî na lêgë ti technique — AILP awärä yëngö, me alîngbi tî fa lïngbïngö la pëpe.
  • Lïsî ti undêrê ayeke mbeti-kôzo; angbâti so ayeke na popo ti fängö, fängö-kêtê, na distillation, azo angbâ na lisoro na ndö ni.
  • Hïngängö ti ndïä ti yëngö ti mändängö akangbi na lêgë ti kodoro oko oko, na andïä na ni la pëpe.
  • v0.1.1: AILP(Resource,ContentUse)=Allow does not imply AARS(Actor,Action)=Allow, and does not imply AADP(Principal,Actor,Authority)=Valid — a system exposing both content and tools must evaluate them separately.