Financial AI for Digitally Excluded Populations: A Technical Approach to Inclusive Finance

Background and objectives

Why inclusive finance matters

Smartphones and digitization have made financial services more convenient, but for digitally excluded users (the digitally vulnerable) the experience has often become harder. As bank branches close and complex mobile apps and online banking become the norm, people who struggle with digital tools are left behind by financial services.

This research outlines technical approaches that use AI to improve financial access for digitally excluded populations. We study a financial AI system that helps users achieve professional-grade wealth management through voice and conversation—without requiring deep financial knowledge or technical setup.

Who counts as digitally excluded?

Digitally excluded users (the digitally vulnerable) include people such as:

• Seniors: Older adults who are not comfortable with digital devices
• People with disabilities: Those who find complex interfaces difficult because of visual, hearing, or other impairments
• Foreign nationals: People with limited Korean who struggle with financial terms and procedures
• Marriage migrants: People unfamiliar with Korea's financial system and how transactions work
• Low-income households: People with limited access to devices or connectivity
• Low digital literacy: People who find complex apps or online banking hard to use

Financial access barriers faced by digitally excluded users

1. Physical access

• Fewer bank branches: Branch closures make in-person visits harder
• ATM friction: Complex ATM screens and flows
• Mobility constraints: Difficulty traveling for seniors and people with disabilities

2. Digital interface barriers

• Complex apps: Dense menus and features in banking apps
• Online banking: Multi-step authentication and security settings
• Small screens: Hard-to-read layouts on phones
• Touch UIs: Difficulty for seniors or people with hand tremor

3. Understanding information

• Financial jargon: Limited familiarity with professional terms and concepts
• Technical indicators: Difficulty interpreting RSI, MACD, and similar signals
• Contracts: Long, dense terms and conditions
• Language barriers: Limited Korean proficiency for foreign nationals and marriage migrants

4. Unfamiliarity with how financial transactions work in Korea

• Transfers: Not knowing options such as internet banking, mobile apps, ATMs, or branch counters
• Deposits and installment savings: How to apply and what documents are needed
• Loans: Loan types and application steps
• Tax filing: Deadlines and filing methods
• Household budgeting: How to track income and expenses

How financial AI can help

1. Voice- and text-based interfaces

Using speech-to-text (STT) and text-to-speech (TTS), users can access financial services through natural conversation instead of navigating complex app screens.

• Everyday language: Requests like "Help me manage this safely" or "I want to open a deposit"
• Voice input: Easier control when touch is difficult
• Spoken guidance: Results and instructions delivered by voice—supporting users with visual impairments
• Conversational UI: Text chat for users who prefer typing

2. Step-by-step guidance for financial procedures in Korea

Through voice or text conversation, the system explains how financial transactions work in Korea, step by step. Even when users cannot use complex banking apps or online banking—foreign nationals, marriage migrants, seniors, and others—they can ask questions in plain language and receive clear, easy-to-follow guidance.

3. Multilingual support

Multilingual capabilities reduce language barriers:

• Multilingual dialogue: Conversation in English, Chinese, Vietnamese, and other languages
• Explaining jargon: Translating complex financial terms into the user's language
• Cultural context: Guidance that respects different financial cultures

4. Accessibility

Stronger accessibility helps people with disabilities:

• Visual impairments: Spoken guidance and screen reader support
• Hearing impairments: Text chat and visual feedback
• Motor impairments: Voice control without precise touch
• Cognitive support: Simple, explicit language

5. Plain-language explanations

The AI turns complex financial information into clear, natural-language explanations:

• Indicators: RSI and MACD described as overbought/oversold conditions and trend direction
• Portfolios: Asset allocation explained as "safer" vs. "growth" buckets
• Risk: Complex metrics reframed as loss likelihood and volatility
• Step-by-step flows: Breaking procedures into small, manageable steps

Technical implementation

1. Voice interface (STT/TTS)

Speech recognition and synthesis enable a voice-first experience:

• STT (speech-to-text): Converting user speech to text
• TTS (text-to-speech): Turning AI responses into spoken guidance
• Multilingual ASR: Supporting speech in multiple languages
• Accent and dialect: Handling regional variation where feasible

2. Conversational AI assistant

LLM-based dialogue supports natural, contextual conversation:

• Context: Understanding ongoing dialogue to respond appropriately
• Question understanding: Interpreting varied phrasing
• Guided steps: Walking users through complex procedures
• Confirmation: Checks before consequential actions

3. Knowledge base for financial procedures

A structured knowledge base covers how financial transactions work in Korea:

• Guides by transaction type: Transfers, deposits/savings, loans, tax filing, and more
• Channel-specific guidance: Internet banking, mobile apps, ATMs, branches
• Documents: What to prepare for each case
• Safety: Fraud prevention and security reminders

4. Accessibility engineering

Technical measures for accessibility include:

• Screen reader compatibility: For blind and low-vision users
• High contrast: Easier reading for some visual conditions
• Large text: Options for seniors
• Simple mode: Hiding advanced features to surface essentials

Social value and expected impact

1. Better financial access

Digitally excluded users can reach financial services more easily:

• Physical barriers: Services without always visiting a branch
• Digital barriers: Services without mastering complex apps
• Language barriers: Services without fluent Korean
• Information gaps: Professional-grade explanations for everyone

2. Broader financial inclusion

Inclusive finance means more people can use financial services on fair terms:

• Including marginalized groups: Seniors, people with disabilities, foreign nationals, marriage migrants, and others
• Financial literacy: Clear explanations of terms and procedures
• Wealth management capability: Tools that feel closer to professional advice
• Independence: Less reliance on others for basic financial tasks

3. Narrowing the digital divide

Closing the digital divide reduces information exclusion:

• Information access: Easier access without deep technical skill
• Skill gaps: Advanced features usable without expert knowledge
• Economic opportunity: Better access to financial information expands options

Government R&D and investor perspective

Financial AI for inclusive finance carries significant value from government R&D programs and investor perspectives:

• Social impact: Improving access for digitally excluded users
• Policy alignment: Supporting inclusive finance policy goals
• Market expansion: Reaching users previously underserved
• Technology leadership: Leadership in voice- and conversation-based financial AI
• Transferability: Patterns applicable to healthcare, welfare, and other domains

Future research directions

1. Advancing voice interfaces

• Multilingual ASR accuracy: Better recognition across languages and dialects
• Emotion-aware speech: Adapting responses to user affect
• Voice biometrics: Authentication via voice

2. Personalized guidance

• Tailored explanations: Matched to skill level and preferences
• Learning progress: Gradual onboarding as users improve
• Concrete examples: Scenarios tied to the user's situation

3. Multi-channel delivery

• Smartphones: Voice and text chat in mobile apps
• Smart speakers: Hands-free voice interaction
• Robotics: Personalized robots for guided financial conversations
• Telephony: Voice service over the phone

Conclusion

Financial AI for digitally excluded populations is an important technical path toward realizing inclusive finance. Through voice and conversation, we can help people manage assets at a professional level without requiring deep financial expertise or technical configuration—and move closer to a world where everyone can use financial services with confidence.

In particular, step-by-step guidance for financial procedures in Korea allows foreign nationals, marriage migrants, seniors, and other digitally excluded users— even when they cannot use complex banking apps or online banking—to ask questions in plain language and receive friendly, understandable explanations.

This is not only a technical innovation; it is technology that creates social value. By narrowing the digital divide and expanding financial inclusion, it can contribute to an inclusive financial society where services are accessible to all.