Skip to content

GRA · Teknologi finansial dan transformasi

Keuangan spasialTempat modal bertemu bahaya, keterpaparan, dan transisi

AI dan otomatisasi

Model statistik dan pembelajaran mesin kini menjadi dasar keputusan kredit, underwriting, deteksi penipuan, perdagangan, dan operasi. The Global Risks Alliance (GRA) meneliti kendali yang diperlukan: validasi, ketidakpastian, keterjelasan, pengawasan manusia, dan pengelolaan model pihak ketiga. GRA juga mengajukan pertanyaan sistemik: apa yang terjadi ketika banyak lembaga bergantung pada model dan data yang serupa pada saat yang sama? GRA tidak membangun, menjual, mengaudit, atau menyetujui model apa pun. Bagian dari teknologi dan transformasi keuangan.

02Di mana model digunakan

Otomatisasi di seluruh rantai nilai keuangan.

Kredit

Lending decisions

Scoring, limit setting and early-warning models, where errors affect access to credit and fairness as well as losses.

Asuransi

Underwriting and claims

Seleksi risiko, masukan penetapan harga, dan triase klaim, yang dapat menyesatkan akibat penyimpangan model setelah iklim atau pasar berubah.

Operasi

Penipuan dan operasi

Transaction monitoring, document processing and customer service, where automation scales both efficiency and failure.

03Aturan

AI di sektor keuangan: apa yang berlaku, dan kapan.

InstrumenStatus, Oktober 2026
EU AI Act high-risk duties for creditworthiness and life and health insurance pricingBerlaku mulai 2 Desember 2027 berdasarkan AI Omnibus (Regulation 2026/1744)
Panduan manajemen model AS yang direvisi (SR 26-2)Issued April 2026; generative and agentic AI outside its scope
PRA SS1/23Berlaku bagi bank di Inggris Raya yang memperoleh persetujuan model
Kerangka Manajemen Risiko AI NISTVoluntary
04Observasi Bumi

Pembelajaran mesin pada data satelit, untuk keterpaparan.

One growing use of AI in risk finance is extracting exposure and damage from imagery: building footprints, flood extent from Sentinel-1 radar, crop condition. These models change exposure layers quickly after an event and carry their own error rates, which must travel with the layer. Methods at GCRI disaster monitoring.

05Professional controls

What sound model governance asks.

  1. 01

    Validasi

    Independent testing of a model’s design, data and performance against its intended use, before and during deployment.

  2. 02

    Ketidakpastian

    Quantifying how wrong a model could be and under which conditions, using methods such as those in uncertainty quantification and model validation.

  3. 03

    Keterjelasan

    Mampu menyatakan mengapa suatu model menghasilkan keluaran tertentu, kepada orang yang terdampak dan kepada pihak yang mengawasinya.

  4. 04

    Pengawasan manusia

    Titik-titik yang jelas tempat orang yang kompeten dapat meninjau, mengesampingkan, atau menghentikan keputusan otomatis.

  5. 05

    Pihak ketiga

    Mengetahui isi model vendor, bagaimana model itu berubah, dan apa yang terjadi jika vendor gagal.

06Systemic view

Ketika semua pihak menggunakan model yang sama.

Model risk is usually managed institution by institution. GRA studies the shared case. Common vendors, training data and methods can correlate errors, so many institutions misjudge the same risk at the same moment.

  • Which models and data sources are widely shared across a sector?
  • How would a common error show up in credit, claims or market behaviour?
  • What signals would reveal correlated model failure early?
  • Which questions belong to supervisors, auditors or each institution’s own validation function?
07Batasan dan keterhubungan

Research, not assurance.

GRA validates, audits, rates and certifies no model or vendor. Its research is not a compliance opinion. Each institution stays accountable for its own models. The risk channel: AI and model risk. A proposed AI and Financial Model Risk community is an established platform. Related lines: digital assets research and quantum and cryptographic transition. The Global Risks Forum (GRF) discusses AI in public decisions on digital and AI governance.

09Pertanyaan

Tanya jawab.

Does GRA use AI in its own research?
Where GRA uses automated methods in research, the publication states the method and its limits, and findings are reviewed by a person before release.
Can GRA review my institution’s model?
No. GRA does not perform any institution’s own model validation or risk-management work. Its outputs are industry-wide research.
Is this related to the GRA platform Fintech Nexus?
Yes. Fintech Nexus covers AI in financial services as part of its sector community; this page frames the cross-cutting research questions.
How can practitioners contribute?
Model validators, data scientists and risk managers can register interest at the membership application form. Working groups open when their terms are adopted.
LaporanBerkontribusi

Kontribusikan bukti dari halaman ini.

Reports adalah publikasi triwulanan Konsorsium. Apa yang Anda ketahui tentang subjek ini dapat menjadi bukti, kasus, atau koreksi dalam edisi berikutnya, dengan kredit atas nama Anda dan dipublikasikan di bawah CC BY 4.0 jika merupakan karya Anda sendiri.

Global Risks Alliance
The Global Risks Alliance

Ajukan pertanyaan eksposur atau modal kepada Aliansi.

Informasi dan pertanyaan sudah terbuka hari ini. Sebuah pertanyaan menerima nomor referensi dan menyebutkan langkah berikutnya; lamaran, pembayaran, alokasi komputasi, dan pendaftaran hanya dibuka ketika tim, ketentuan, dan kapasitas tinjauannya tersedia.