AILP — Protokol Kebenaran Pembelajaran AI

01 · Definisi

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 · Tujuan

• Jawab "bolehkah AI belajar daripada ini?" dengan ketepatan peringkat dimensi dan bukan sekadar bit rangkak.
• Pisahkan kebenaran pembelajaran (penginternalisasian ke dalam model) daripada kebenaran akses (mendapatkan halaman).
• Menyokong model pampasan: percuma, bukan komersial, memerlukan lesen, perkongsian hasil.
• Pastikan keterbukaan boleh dilaksanakan mesin — supaya niat baik tidak dibuang sebagai ketidakpastian undang-undang.

03 · Skop

04 · Contoh boleh baca mesin

Deklarasi kebenaran pembelajaran — /ai/rights-spectrum.json (profil AILP)

{
  "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 · Batasan

• Semantik kebenaran pembelajaran adalah yang paling sukar disahkan secara teknikal — AILP mengisytiharkan niat, ia belum lagi dapat membuktikan pematuhan.
• Senarai dimensi ini adalah draf; sempadan antara latihan, penalaan-halus, dan penyulingan masih diperdebatkan.
• Pengiktirafan undang-undang terhadap kebenaran pembelajaran berbeza mengikut bidang kuasa dan masih belum diselesaikan.
• 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.