FINANCIAL AI
Financial markets,
through a technical lens.
Articles from our research and validation work, covering financial time series, machine-learning models, signal detection, and AI agent design.
Articles in this category are not recommendations to buy or sell any financial product.
The FSB's Final Report on 12 Sound Practices: A Materiality and Risk Perspective
Materiality and risk assessments help institutions tailor controls to each AI use case. We examine how Sound Practice 5 connects those assessments to documentation, validation and human oversight.
FINRA's Six AI Agent Risks and Supervisory Procedures
FINRA's six risks and its considerations, read apart from the obligations in Rule 3110—and carried through to what a written supervisory procedure actually has to say.
ASIC's Focus on Banks' Use of AI: What Records to Keep for Customer-Facing AI
ASIC's corporate plan devotes essentially one item to banks and AI. We check what it does and does not say against the primary sources, and turn that into evidence design for customer-facing AI.
Mercury Agent Cards: Enforced Spend Limits and Customer Liability
Telling an agent "spend up to this much" in a prompt is not the same as a limit that gets declined at authorization. And setting a limit does not move liability.
EU AI Act Becomes Generally Applicable: Key Annex III High-Risk Obligations Deferred to 2027
The AI Act became generally applicable on 2 August 2026, and the transparency obligations now apply. With key Annex III high-risk obligations deferred to December 2027, here is how the two timelines separate.
The FSB's Consultation on AI Governance: Human Oversight Takes More Than One Form
"Human oversight" is not one step. Reading the six forms in the FSB's consultation by when each one actually operates.
Before AI Agents Trade on Your Behalf: Verifying Delegated Authority in Finance
Letting an AI trade for you turns less on model performance than on verifiable delegation: identity, scope, and execution—and where each control is enforced.
Machine Learning on Financial Time Series: Built Not to Break
Can you trust a good backtest? A robustness-first way to design validation when applying machine learning to financial time series.