AILP — Porotokole y'Uruhushya rwo Kwiga rw'Ikoranabwenge

Porotokole y'Uruhushya rwo Kwiga rw'Ikoranabwenge

Ikoranabwenge gishobora kwiga kuri ico kintu? Ku bujimbu ki, ku bikorwa ki, hamwe n'inshingano izihe?

01 · Insobanuro

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 · Intumbero

Tanga inyishu ku kibazo "ikoranabwenge coba kibasha kwiga iki?" ku bujimbu bugaragara bw'umugereka, aho gukoresha ikimenyetso kimwe c'ugusakura.
Gutandukanya uruhushya rwo kwiga (ukwinjizwa muri moderi) n'uruhushya rwo kuronka (gufata amapaji).
Gushigikira imfano z'ingurane: ku buntu, atari mu bucuruzi, bisaba uruhushya, gusangira inyungu.
Kwizigira ko ubugaragare bushirwa mu ngiro n'imashini — kugira ngo ubugwaneza butazokurwaho nk'akaga k'amategeko.

03 · Ubugari

kuronka / gushiraho urutonde — kuronka no gushiraho urutonde rw'insobanuro

inference_input — ikoreshwa nk'ibisobanuro mu gihe co gukora ivyavuye (RAG)

embedding — gukora ivyuma vy'imibare no gushiraho urutonde rwo kuronka

training / fine_tuning / distillation — ibipimo vy'ukwinjizwa muri moderi

verbatim_memory — nimba kwigana neza uko byanditswe kwemewe

kwerekana inkomoko / ingurane — inshingano zo gucira n'amabwiriza yo kwishura

04 · Akarorero gasomwa n'imashini

Itangazo ry'uruhushya rwo kwiga — /ai/rights-spectrum.json (umwidondoro wa 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 · Imbibe

Insobanuro y'uruhushya rwo kwiga ni zo zigoye kugenzura mu buryo bw'ikoranabuhanga — AILP imenyekanisha ico bagamije, ariko ntishobora ubu kwerekana ukurikizwa kwaco.
Urutonde rw'imigereka ni inyandiko itarahera; imbibe hagati y'ugutoza, gutunganya birambuye, na distillation ziracari mu mpaka.
Ukwemerwa mu mategeko kw'uruhushya rwo kwiga birahindagurika bivanye n'igihugu kandi ntibiratunganywa.
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.