Artificial Intelligence (AI) in Financial Services Leveraging AI to Serve the Underbanked Steve Njenga, CIO Barclays
• The Underbanked Landscape • What is AI • The Opportunity • FSI/FinTech Partnership Agenda • FinTech Use Cases for AI • Real World Examples • Risks & Challenges
The “Invisible” Population The Underbanked Who are they?? The numbers…… +365,000,000 in Africa Landscape Potential: 1.6 billion new retail customers globally Proliferation of Data Processing power increasing exponentially Low-cost Android smartphone 5.7 billion mobile phone suscriptions (2016) Six in 10 global smartphone connections will come from developing regions within five years 35 billion terabytes of data by 2020
AI is the theory and development of computer systems that can perform tasks that normally require human intelligence Artificial Intelligence visual perception speech recognition decision-making translation AI data analysis (IoT, social media, mobile phones etc.)
FSI Goal …… The Argument……. ……. increase The Opportunity profits by Too Risky increasing Financial Services penetration Unprofitable However……. ……. diminishes the challenges Regulatory Environment…… of profitability in serving the “invisible” population Digital Technology……
+ve -ve FSI/FinTech Partnership • Niche focus • Reg. constraints • Skills • Cash strapped • Agility • Trust • Quick to market Partnership Instead Of Competition +ve -ve • Plenty of cash • Slow, lethargic • Min Reg. constraints • Bureaucratic • Customer base • Skills • Trust & relationships
……AI powered services FinTech Use Cases Visual Identification & Verification • Leveraging capsule neural networks technology to visually identify customers and documents Advanced “Big Data” Analytics • Examining data or content using sophisticated AI techniques and tools (beyond those of traditional BI), to discover deeper insights, make predictions, or generate recommendations. Intelligent Virtual Assistants • Emulating human interaction: AI powered chat-bots that can intelligently answer customer queries, effectively reducing load from customer services department.
• Credit rating via AI to make loans in emerging markets Real World Examples • Combining non-traditional data sources and machine learning • Paperless and fast application processes with immediate scoring and pay-out • Credit scoring and fraud detection using AI • Provide loans, insurance and transactional banking • WhatsApp lending - instant credit by starting up a conversation with a chatbot • Haraka : nano loans paid in real-time into your mobile wallet .
Real World Examples • Credit assessment algorithms that collect and analyze psychometric and behavioral information • Measure something that all borrowers have • character, their abilities, and willingness to repay. • Deeper and more quantitative understanding of risk by gauging ethics, honesty, intelligence, attitudes and beliefs . https://producttour.eflglobal.com/ • Uses AI and Deep Learning to provide you with Virtual Assistants • True conversational language solutions • Voice Biometrics driven through conversational language interactions
Legacy systems Risk & Challenges Controls: ‘ Do we have the internal controls? ’ and ‘ Do our partners have them? ’ Operational complexity & Cost Partnering with third-party tech players opens up new vulnerabilities New Technology = additional vectors for hacking and accessing funds
Recommend
More recommend
