766 deep dives in 24 categories · last updated
16 matching deep dives
Research Lab: these deep dives are AI working drafts unless tagged Reviewed research. Read how review works.
AI workflows, software and cloud infrastructure, data standards, privacy-preserving methods, cybersecurity and technology controls.
Passkeys resist credential phishing through public-key authentication tied to the genuine service. Enrollment, lost devices, fallback access and payment authorization remain distinct parts of the customer’s experience.
Copilot Studio connects information, agent behavior and external services. Authentication, connector permissions and execution records determine what an agent can actually read or do; conversation volume alone says little about the result.
Spring Labs sells AI software that reviews bank customers' calls, emails and chats to find complaints and check service quality. On October 7, 2026, it acquired Winnow and launched SpringGRC, a compliance-management platform now in limited beta.
AWS offers configurable filters around AI interactions. The important boundaries are which content reaches them, what a grounding score means and how many legitimate requests are blocked alongside harmful ones.
Research shows how institutions can train together while keeping original records in place. Privacy depends on the exchanged updates, record matching and model outputs, while usefulness depends on whose data and outcomes the model represents.
Fiddler describes tools for observing predictive models, tracing AI agents and filtering requests. The distinction between an early warning, a confirmed error and a successful correction explains both their usefulness and their limits.
Retrieval lets an AI system bring filings and policies into an answer. Accuracy still depends on which version it finds, which exceptions survive extraction, and whether each claim follows from the evidence.
Cloud services offer flexible capacity while linking financial firms to shared infrastructure. A second region, a backup and a route to another provider address different failures, with different costs and limits.
Central-bank experiments show how post-quantum signatures can work in payment systems. Moving from a successful test to durable protection involves data lifetimes, software, counterparties and the capacity to process payments.
Generated records can help test financial software and study relationships without distributing original files. A realistic-looking dataset still needs separate evidence about privacy, statistical usefulness and performance on real outcomes.
Fannie Mae and Freddie Mac retain the November 2, 2026 mandate while allowing approved Sellers extra time. The redesign makes appraisal information more structured, but comparable data still requires judgment about comparable homes.
DataRobot’s monitoring tools show how deployed AI models and agents behave. Their value depends on connecting those records to what happened to the customer and the work needed to resolve a problem.
Differential privacy limits what a statistical release can reveal about a defined contribution. The protection brings a measurable accuracy tradeoff, especially for small groups and repeated releases.
Handing checkout to a payment provider can reduce a merchant’s exposure without ending its responsibilities. PCI DSS follows the card data and the systems that can affect its security.
An AI assistant can read a document without giving its author authority to direct the task. Prompt injection exploits that boundary, and its consequences depend on the tools and data the assistant can reach.
XBRL makes reported financial facts easier for software to retrieve. Their periods, definitions and filing history still determine whether two numbers belong in the same comparison.
Page 1 of 1 · 1–16 of 16 articles