AI explainability in banking, explained for examiners, RFPs, and compliance: what SR 11-7, the EU AI Act, and Reg B actually require — and how to build for it.
Enterprise AI & Engineering Insights — Page 4
Showing 37 - 48 of 623 Posts
A sustainability filing is structured data, not prose. XBRL export turns measured, ESRS-E1-mapped AI emissions into a tagged, audit-ready figure.
Finance forecasts AI spend in dollars; almost nobody forecasts it in carbon. As disclosure tightens, the unforecast AI emissions line becomes a surprise.
The same AI workload emits differently depending on where and when it runs — grid-aware measurement uses the real carbon intensity, not a flat average.
Map per-inference AI emissions into ESRS E1 with a defensible methodology — the step that turns a carbon number into a CSRD disclosure.
A design gate passed a blank page. Sound, vacuous or broken — how to tell which of your checks actually catches anything, using known-bad fixtures.
OWASP ranks prompt injection the #1 LLM risk. Learn direct vs indirect injection, why it can't be fully patched, and a defense-in-depth framework for banks.
Compare on-premise, private cloud, and sovereign cloud deployment for a bank's LLM, with a decision framework matched to regulatory and risk profiles.