The Complete Guide to Using AI in the Financial Services Industry in Indonesia in 2025

Indonesia's 2025 AI moment matters for financial services because the government's Indonesia national AI roadmap 2025 names economy and finance as priority sectors, pairs talent and infrastructure targets (including a goal of producing 100,000 AI talents a year) with ethical guardrails, and maps financing options to scale pilots into nation-wide systems; regulators are also putting ethics front-and-center as the roadmap is finalized in late August 2025 (Indonesia AI ethics framework prioritized in August 2025).

For banks and fintech the upside is concrete - smarter credit scoring, fraud detection, and personalized services that can extend reach to underbanked customers - yet gaps in skills, research funding and regional connectivity mean practical reskilling is urgent; programs like Nucamp's Nucamp AI Essentials for Work bootcamp (15-week AI training) are one pathway to build the hands-on skills needed to turn policy into products that really reach millions.

Metric	Target / Timeline
AI talent production	100,000 annually
AI literacy	20 million citizens by 2029
Sovereign AI Fund	Planned 2027–2029 (public‑private model)

“The very first rule will most likely focus on AI ethics.” - Minister Meutya V. Hafid

Indonesia's 2025 AI horizon is headed toward agentic systems that act with real autonomy - streamlining onboarding, automating risk and AML checks, and delivering hyper‑personalized services that can finally scale across 17,504 islands and a smartphone base of over 180 million users; that combination of reach and complexity makes agentic AI especially potent for Indonesian banks and fintechs, but it also raises familiar trade‑offs around data quality, privacy and governance that regulators and industry must solve together.

Evidence from regional and industry reports points to big upside: AI is already boosting fraud detection and tailored credit scoring to pull more customers into the formal system, financial inclusion rose to about 83–84% by 2023, and market forecasts show agentic AI in financial services becoming a material commercial sector (market estimates put the agentic AI market at roughly USD 5.51 billion in 2025).

The practical path forward is clear - prioritize data infrastructure, model orchestration and multilingual agents that can navigate local contexts, while pairing pilots with governance and reskilling so the technology lifts underbanked communities instead of leaving them behind; for a deeper read on Indonesia's inclusion story see the World Economic Forum overview on financial inclusion and for the agentic mechanics and risks see reporting on agentic AI systems in Indonesian finance.

Metric	Value / Year
Population	Over 280 million
Smartphones	Over 180 million
Internet penetration	79% (2024)
Financial inclusion	~83–84% (2023)
Projected GMV (digital economy)	$200–360 billion by 2030
Agentic AI market (FSI)	~USD 5.51 billion (2025 est.)

“Insurance decisions often hinge on timely, accurate and contextually relevant information - especially when it comes to quoting across products like car, home, travel, and small and medium enterprises (SME).” - Augustine Tay, InsureMO

AI in Indonesia's financial services today shows up where customers spend most of their time: chat windows and messaging apps, backed by multilingual models, retrieval‑augmented pipelines and API orchestration that move simple requests from phone to resolution without a human in the loop.

Telkomsel's Veronika handles about 95% of routine queries, PasarPolis's “Poli” manages tens of millions of micro‑insurance policies monthly, and Bank Rakyat Indonesia's generative assistant “Sabrina” answers balance checks, generates mini‑statements and can even trigger card locks or loan pre‑approvals - concrete examples of automation trimming queues and cutting operational cost while improving speed and access.

Homegrown vendors like Kata.ai, Botika, Bahasa.ai and Chatbiz supply voice and WhatsApp‑integrated agents, RAG for enterprise Q&A, and low‑code flow builders that plug into CRMs, payment rails and logistics, helping banks and fintechs scale onboarding, fraud screening and personalized offers across 212 million internet users (75% mobile as of Jan 2025).

Those capabilities are already shifting customer expectations and creating measurable lift in inclusion and efficiency across Indonesia's archipelago; see the market roundup of chatbot vendors and the broader Indonesia AI revolution for details.

Vendor	Key capability
Botika	GPT-powered NLG; Omnibotika for multi-step API orchestration
Kata.ai	Multimodal, multilingual LLMs; voice + chat integration (clients: BRI)
Bahasa.ai	RAG knowledge retrieval; enterprise Q&A and payments/logistics integrations
Chatbiz	Low-code multi-turn flow builder; CRM/ERP connectivity

“Indonesians are not just users of AI, but creators and innovators.” - Vikram Sinha

AI is turning what used to be invisible into bankable: by combining alternative data from smartphones, payment flows and digital footprints with machine‑learning scorecards, lenders can now assess people like Sari, a Jakarta food‑stall owner who lacked formal records but has a steady income, and offer affordable credit that was previously out of reach; concrete pilots show this works - GBG's partnership with CredoLab uses mobile metadata to lift scorecard predictiveness by up to ~40%, cut cost of risk and raise approval rates by as much as 32% - proof that fraud detection and risk models can move from exclusion to inclusion (GBG CredoLab AI-driven risk management partnership in Indonesia).

Evidence syntheses and field studies reinforce the promise - and the caveats - of alternative‑data scoring: machine learning can outpredict traditional credit models but also brings integration, bias and data‑governance challenges that require careful policy and product design (J‑PAL review on alternative data and AI for financial inclusion).

At a national level the World Economic Forum notes how Indonesia's vast smartphone base, rising internet penetration and supportive blueprints are creating the conditions for AI to expand access - if regulators, industry and civil society coordinate on privacy, fairness and interoperability (World Economic Forum analysis on AI adoption in Indonesia).

The takeaway is practical: deploy alternative‑data models alongside strong governance, transparency and reskilling so AI actually broadens opportunity instead of amplifying existing gaps.

Metric	Key figure (from sources)
Smartphone users	Over 180 million
Internet penetration	79% (2024)
Financial inclusion index	~84%
Target unbanked/new-to-credit reach cited	181 million (GBG)
Model uplift reported	~40% predictiveness; approval ↑ up to 32%

“This convergence between mobile credit scoring and digital fraud technology helps established and digital banks and lenders onboard quality customers within the financially excluded population.” - June Lee, Managing Director APAC (GBG)

OJK's April 29, 2025 launch of

Artificial Intelligence Governance for Indonesian Banks

sets a clear baseline for responsible AI in the country's banking sector: it covers advanced AI systems, expects banks to embed risk management, human oversight and prudence, and is explicitly designed to be the

minimal benchmark

that complements existing rules like the Blueprint of Banking Digital Transformation and POJK 11/POJK.03/2022 on IT implementation; in short, this guidance is the regulatory safety net under increasingly autonomous agents that banks are deploying for credit, fraud detection and customer service.

The framework aligns Indonesian practice with international norms while remaining rooted in local law - Dentons' practical overview notes alignment with the Personal Data Protection Law and highlights the need for explainability, ethical standards and operational controls - so banks turning pilots into production must hardwire governance, documentation and compliance from day one.

For banks and vendors, the message is practical and urgent: treat the guidance as the new baseline for design, testing and vendor oversight so AI systems expand access safely instead of amplifying risk.

Attachment	Notes
OJK Artificial Intelligence Governance for Banking - Attachment 1 (PDF)	Main governance document (April 29, 2025)
Lampiran 2	Supporting annex
Lampiran 3	Supporting annex
Lampiran 4	Supporting annex
Lampiran 5	Supporting annex

Indonesia's national AI roadmap turns ambition into a concrete playbook for finance: the White Paper (building on the 2020–2045 Stranas KA) maps short‑, medium‑ and long‑term actions that pair talent goals - produce 100,000 AI specialists and make 20 million citizens AI‑literate by 2029 - with targeted infrastructure and financing to scale real products in economy and finance; the plan calls for expanded high‑performance computing, GPUs/TPUs, a sovereign national cloud and green data centres via public–private partnerships, and even a Danantara‑led Sovereign AI Fund and blended financing to move pilots into nationwide systems (see the published Indonesia national AI roadmap and White Paper and the OECD summary of Stranas KA).

Global and local investors are already responding with large cloud and GPU commitments and regional centres of excellence that will help banks and fintechs access the compute and language models they need; practical implications for finance are direct - secure, sovereign compute and predictable blended financing make it possible to run privacy‑preserving scorecards, national sandboxes and multilingual agents at scale while tying investment to governance and reskilling (for an infrastructure and investment perspective see reporting on Indonesia's AI investment momentum).

The takeaway is pragmatic: the roadmap aligns policy, cash and compute so that AI can reach clients across thousands of islands - provided regulators, industry and funders follow through on the staged financing and infrastructure targets.

Metric / Target	Detail
AI talent production	100,000 annually
AI literacy	20 million citizens by 2029
Core infra	HPC, GPUs/TPUs, sovereign national cloud, green data centres
Financing vehicle	Danantara / Sovereign AI Fund; blended public‑private financing

“Whichever country controls AI can potentially control the world.” - President Joko Widodo

The AI industry outlook for 2025 in Indonesia is one of rapid scaling and high-stakes opportunity: strong government roadmaps and flagship investments are turning ambition into infrastructure (Microsoft's record $1.7B cloud commitment and new GPU centres), local momentum is building around language models and sovereign clouds, and finance sits squarely in the fast lane where AI already drives fraud detection, credit scoring and personalized services; market trackers put Indonesia's AI market in a multi‑billion dollar trajectory - homegrown estimates show the national AI market rising from about USD 2.4B today toward roughly USD 10.88B by 2030 while broader digital GMV could reach $200–360B by 2030 - so banks, fintechs and vendors that secure compute, compliance and multilingual models can capture outsized returns.

Structural bets are clear: expand AI‑optimised data centres, deepen generative AI adoption in banking and insurance, and pair deployments with reskilling and governance to close the gap between urban pilots and archipelago‑wide services; for a policy‑and‑inclusion view see the World Economic Forum analysis and for infrastructure and market forecasts see the Indonesia AI Revolution briefing.

Metric	Figure / Year
Current AI market (reported)	~USD 2.4B (Introl)
Projected AI market	~USD 10.88B by 2030 (Introl)
Projected digital GMV	USD 200–360B by 2030 (WEF)
AI‑optimised data centre market	USD 0.66B (2025) → USD 1.44B (2030) (Mordor)

“Indonesians are not just users of AI, but creators and innovators.” - Vikram Sinha

Building AI talent for Indonesia's finance sector must be a two‑track effort: create a steady pipeline from schools while rapidly reskilling existing bank and fintech teams.

The national move to offer elective AI and coding in primary and secondary schools (rolling out in 2025–26) - with classrooms earmarked for two hours a week in fifth grade and more time at higher grades - lays the long‑term foundation, but practical readiness hinges on teachers, digital infrastructure and context‑aware curricula (GovInsider: Indonesia AI classroom rollout and teacher training).

Experts at UGM stress that ethics, logic and digital literacy must come first so students learn critical thinking, not just tools (UGM: staged ethical AI learning and strengthening digital literacy).

For the finance industry this means partnering now with schools, vocational programs and private trainers to co‑design modules (including “unplugged” options for remote regions and satellite internet plans) and to fund bootcamps and on‑the‑job reskilling that teach model ops, data governance and customer‑facing AI skills; think of it as stitching a national classroom into an industry apprenticeship pipeline so that credit scoring, fraud detection and customer‑service agents are built by locally trained teams, not just imported talent.

“In practice, the material needs to be delivered in stages. We cannot just start teaching AI applications to elementary school children; that would be a disaster. We must first equip them with logic, ethics, and digital literacy.” - Iradat Wirid, CfDS UGM

As AI scales across Indonesia's 17,504 islands, the upside for inclusion - smarter risk models, faster fraud detection and personalized services - is matched by concrete hazards that must be governed: data privacy, model bias, cyberattacks and systemic operational fragility if agents fail or are abused.

Regulators and industry are already responding with guidance and blueprints, but real resilience requires banks to pair innovation with hardened cyber maturity, clear accountability and continuous testing; one empirical study found a cybersecurity maturity score of 2.1 with an inherent risk score of 1.9, a reminder that current controls work but need steady improvement to keep pace with evolving threats (Indonesia cybersecurity maturity study).

Policymakers also stress a balanced approach - protect consumers and data sovereignty while enabling pilots that improve access - so layered safeguards (privacy-preserving scorecards, explainability requirements, incident playbooks and vendor oversight) should be non‑negotiable parts of production pipelines.

Think of regulation and resilience as a national safety net: without it, a single large fraud or outage could ripple across the digital economy; with it, AI can safely extend services to millions - an outcome the World Economic Forum says hinges on coordinated policy, industry ethics and robust cyber defenses (World Economic Forum report on the rise of AI in Indonesia).

Risk / Metric	Figure / Note
Population / geography	Over 280 million; 17,504 islands
Smartphone users	Over 180 million
Internet penetration	79% (2024)
Financial inclusion index	~84% (2023)
Projected digital GMV	USD $200–360B by 2030
Cybersecurity maturity (study)	Maturity score 2.1; inherent risk 1.9

Conclusion and a beginner's action checklist for adopting AI in Indonesia: the path is practical - start small, align with rules, and scale with skills and infrastructure.

Begin by learning usable tools and prompt skills (a 15‑week course like Nucamp AI Essentials for Work bootcamp (15‑week AI at Work program) teaches hands‑on prompts and workplace AI use), because 90% of workers say they want training and a young, tech‑savvy population already drives rapid adoption; next, map regulatory must‑dos - use guides such as the Fintech laws and regulations in Indonesia 2025 (Global Legal Insights) to check sandbox pathways, SLIK reporting and OJK/BI obligations before any pilot; run a small privacy‑preserving pilot in a regulatory sandbox, measure uplift on inclusion and fraud reduction, then hardwire governance, vendor oversight and cybersecurity as you scale; finally, factor in infrastructure and financing - Public First's AI Opportunity Report for Indonesia (Public First) highlights compute, connectivity and public‑sector efficiency gains as levers to expand impact across Indonesia's islands.

Think of this checklist as a starter map: learn, comply, pilot, secure, and partner - so AI grows access safely instead of increasing risk.

Beginner Action	Why / Source
Practical training (15 weeks)	Nucamp AI Essentials for Work: hands‑on prompts & workplace AI (Nucamp AI Essentials for Work syllabus (15‑week course))
Regulatory checklist	Review GLI OJK/BI guidance, sandbox rules and SLIK reporting (Global Legal Insights - Fintech laws in Indonesia)
Start a sandbox pilot	Test privacy‑preserving scorecards and fraud models before production (OJK/BI sandbox pathways)
Plan infra & partnerships	Public First: invest in compute, connectivity and reskilling to seize national AI opportunity (Public First AI Opportunity report - Indonesia)