AILP — Ubwiriza bw'Uruhushya rwo Kwiga rwa AI

Ubwiriza bw'Uruhushya rwo Kwiga rwa AI

AI ishobora kwiga kuri ibi? Ku rwego rungana iki, ku bikorwa ki, hamwe n'ibisabwa ki?

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 · Intego

Gusubiza ikibazo "AI ishobora kwiga kuri ibi?" mu buryo burambuye ku rwego rwa buri gice, aho gukoresha akamenyetso kamwe gusa.
Gutandukanya uruhushya rwo kwiga (kwinjiza mu ngiro za moderi) n'uruhushya bwo kubona (gutangira amapaji).
Gushyigikira uburyo bw'indishyi: ku buntu, bitari ubucuruzi, bisaba uruhushya, gusangira inyungu.
Kwemeza ko ubwisanzure bushobora gukurikizwa na mudasobwa — kugira ngo ubushake bwiza budasuzumwa nk'ikibazo cy'amategeko kitazwi neza.

03 · Urwego rukubiyemo

access / indexing — gutangira no kwandika mu bubiko hakurikijwe insobanuro

inference_input — ikoreshwa nk'ibirimo mu gihe cyo gusesengura (RAG)

embedding — guhindura mu mibare no kwandika mu bubiko bwo kubona

training / fine_tuning / distillation — inzego zo kwinjiza ubumenyi muri moderi

verbatim_memory — niba kwisubiramo mu buryo buhwanye n'inyandiko y'umwimerere byemewe

attribution / compensation — inshingano zo kwerekana inkomoko n'ibisabwa byo kwishyura

04 · Urugero rusomwa na mudasobwa

Itangazo ry'uruhushya rwo kwiga — /ai/rights-spectrum.json (imiterere ya 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 · Imbibi

Insobanuro z'uruhushya rwo kwiga ni zo zigoye cyane kugenzura mu buryo bwa tekiniki — AILP itangaza intego, ariko ntishobora kubwo gukomeza kwerekana ubwubahirizacyo.
Urutonde rw'ibice ni umushinga; imbibi hagati yo kwitoza, gutunganya neza, na distillation zikiri mu mpaka.
Kwemera mu mategeko uruhushya rwo kwiga bitandukanye bitewe n'ubutegetsi bw'amategeko kandi ntibirasobanurwa neza.
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.