FinTech - Shaping the Financial World mit course finance technology fintech

Artificial Intelligence in Finance


  • Fintech stack
  • AI and machine learning in finance
  • Finance public policy framework
  • AI policy and finance
  • VISA attempted to acquire Plaid for ~$5 billion
    • Deal blocked due to anti-trust

Financial Technology Stack

  • TradFi (money, accounting, ledgers, markets, etc.)
    • Internet
      • Mobile
        • Cloud
          • AI / machine learning / NLP
          • Open API
            • Blockchain? --> not quite in the stack yet

AI and Machine Learning

  • Terms
    • AI
      • 1950s
      • Computers mimicking human intelligence
    • Machine learning
      • 1980s
      • Machines improve with experience
    • Representation learning
      • Machine learning extracting features of data sets
    • Deep learning
      • 2010s
      • Machine learning with multi-layer neural networks
  • AI as a service or a tool?
    • Typically thought of as a tool
    • There are plenty of examples of AI as a service (in finance):
      • AlphaSense - search engine
      • Cape Analytics - insurance and property risk analytics
      • Dataminr - market sentiment analysis
      • HyperScience - document processing
  • Big banks/finance are working to integrate this into their businesses
    • Third party + roll their own
    • Lots of jobs for big AI/ML adjacent positions (from 2019):
      • Chief data scientist at Goldman Sachs
      • Machine learning manager at J.P. Morgan
      • AI research manager at Morgan Stanley
      • Senior AI technology architect at Bank of America

AI Policy and Finance

  • When working with AI, it's always important to ask: is there bias?
  • Who owns data?
    • Banks? the third party using it for ML? The users?
  • Use of alternative data
    • Not just bank transactions but driving records, social media, etc.
  • How do we fit AI into the current frameworks?

Policy Alternatives

  • Options for moving forward:
    • Be tech neutral
    • Adjust regulatory requirements
    • Promote early stage tech


Study Questions / Issues to Prepare

  1. How do new forms of AI enabled data analytics, pattern recognition, chatbots, natural language processing, and robotic process automation fit within other emerging FinTech trends?
    • Our ability to store and process data
  2. How has it affected the competitive landscape of financial incumbents, big tech firms, and FinTech startups?
  3. What new public policy challenges are AI and machine learning presenting for the fairness, explainability, privacy, robustness, and resiliency of the financial sector? How do and should these new applications and tools fit within the regulatory structure of the financial industry?
    • Need to ensure certain biases (race, gender, etc.) in models don't exist - Black box reasoning isn't acceptable