Real people, real budgets, and real financial decisions broken down step by step so you can see how financial ideas work beyond the classroom.
Payment networks and banks such as Mastercard (HQ: Purchase, New York) deploy machine-learning fraud scoring that protects all customers, including students, from account takeover and scam payments, while false-positive handling remains a key challenge.
US student loan servicing increasingly applies machine learning to predict borrower repayment difficulty, route support, and personalize outreach; Sallie Mae (HQ: Newark, Delaware) discloses analytics use, though public detail on specific AI models is limited.
Cleo, a UK-founded fintech headquartered in London, offers an AI chatbot that connects to users' bank accounts via open banking, analyzes spending, and delivers budgeting nudges through conversational messaging aimed at students and young adults.
Morgan Stanley deployed an OpenAI-powered assistant (AI @ Morgan Stanley Assistant) in 2023, in partnership with OpenAI, to help ~16,000 financial advisors search internal research, summarize documents, and prepare client-facing material. This case examines capabilities, human oversight, phased rollout, and reported adoption.
Kyriba, a cloud treasury management provider headquartered in San Diego, embeds AI-based cash-flow forecasting and ML-driven cash prediction within its treasury platform, enabling CFOs and treasurers to improve liquidity visibility, forecast accuracy, and cash planning using their own historical data.
JPMorgan Chase developed COIN (Contract Intelligence), an AI platform that automates review of commercial credit agreements and legal documents, famously reported as saving roughly 360,000 hours of annual lawyer and loan-officer work. This case examines its design, human oversight, and limits of public evidence.
Morgan Stanley deployed GPT-4-based AI tools to help its financial advisors search and synthesize internal research, market commentary, and firm knowledge. The tools aim to reduce research retrieval time and let advisors focus on client relationships, with human advisors retaining full decision-making authority.
JPMorgan Chase developed its Contract Intelligence (COIN) platform to automate the review of commercial loan agreements, work previously done manually by lawyers and loan officers. The platform extracts key data points and flags contract deviations, dramatically reducing review time and document processing errors.
HSBC partnered with Google Cloud to apply AI to transaction monitoring for financial crime detection. The system identifies suspicious activity patterns that rule-based systems miss, cutting false positives and enabling investigators to prioritize genuinely high-risk cases across HSBC's global transaction flows.
Tala, a Santa Monica-based fintech, uses smartphone data and machine learning to underwrite microloans for consumers without formal credit histories in Kenya, the Philippines, Mexico, and India. The company reports millions of customers served and approval decisions in minutes, though most outcome metrics are self-reported.
PayPal applies deep learning and large-scale machine learning across its payments network to detect account takeover, identity theft, and fraudulent transactions in real time across hundreds of millions of accounts. Published company statements report maintaining fraud loss rates well below industry averages, though precise current figures are proprietary.
JPMorgan Chase deployed COiN, a machine learning platform for reviewing commercial loan agreements and credit documentation, replacing manual legal review. Publicly reported figures indicate roughly 360,000 hours of annual lawyer/loan-officer review work automated, with fewer errors, though detailed audited metrics remain limited.
In a seven-day live pilot, Visa and Lloyds Banking Group settled US$750,000 in payment obligations using USDC stablecoins purchased through Archax, a UK-regulated digital-asset exchange. Funds reached Visa in under an hour, including over a weekend, testing whether 24/7 settlement can relieve liquidity lockups inherent in traditional correspondent banking windows.
Sun Finance, a Latvia-headquartered consumer finance group operating across Europe, Latin America and Asia, faced a common scaling problem for digital lenders: verifying customer identities quickly without opening the door to fraud. This case study examines how the group applied AI-driven identity verification, document processing and fraud detection in its lending journeys, drawing on published c
