AI in financial services is not one thing. It is a spreading tool that reads a borrower's tax return and a draft credit memo waiting for a credit officer, a gradient-boosted credit score whose top factors must become accurate adverse-action reasons, a drafted Suspicious Activity Report narrative in an investigator's case queue, a fraud score that declines a card in milliseconds, a summary of last night's earnings call on a research terminal, an AI notetaker's draft follow-up email to a client, a loss-run summary in an underwriting workbench and a photo-based repair estimate on an adjuster's desk. This free hour walks through those jobs one at a time, as the person who does them sees them.
For each job you see how the work is done today, what the AI actually does and with which data, the named products in use as of October 2026 (for example nCino Banking Advisor, Moody's CreditLens, Zest AI, NICE Actimize X-Sight AI Narrate, Quantexa, Mastercard Decision Intelligence, Visa's Featurespace, Feedzai, Stripe Radar, Bloomberg's AI earnings call summaries, AlphaSense, S&P Capital IQ Pro, Jump, Zocks, Sixfold, Federato, Cytora, CCC Estimate-STP and Tractable), where the human decision must stay human, how to measure whether the tool is working, and the specific control that applies: the April 2026 interagency model risk guidance SR 26-2 (aimed mainly at banks over $30 billion in assets) and its carve-out for generative and agentic AI, Regulation B's specific-reasons rule (12 CFR 1002.9(b)(2)), the 30-day SAR deadline and SAR confidentiality (31 CFR 1020.320 for banks, 1022.320 for money services businesses), Regulation E and Regulation Z dispute timelines, Nacha's 2026 fraud-monitoring rules, FINRA Regulatory Notice 24-09, FINRA Rule 2241 and Regulation AC for research, Regulation Best Interest, SEC recordkeeping rules and the SEC's AI-washing cases, Regulation S-P, the NAIC Model Bulletin on insurers' use of AI (adopted in 29 jurisdictions as of October 8, 2026), NY DFS Circular Letter No. 7, and the EU AI Act's Annex III rules for US firms that lend or insure in the EU. Products are named to show what exists, not to endorse them, and vendor performance claims are labeled as claims. Each lesson takes about five minutes to read; code is shown with its output, so running it is optional. Every lesson ends with an optional exercise on fully synthetic data, such as finding two planted errors in an AI-drafted credit analysis or computing precision and recall at three fraud-score thresholds. This course is educational and is not legal or regulatory advice. A 10-question exam earns a free Wirewalk Institute certificate of completion and assessment, and the last lesson maps each role to the paid course that goes deeper.