SR 26-2 Has No Checklist. How to Find the Gaps in Your Model Risk Program
SR 26-2 sets no prescriptive requirements, so there is no official checklist. A free self-assessment rates your model risk program in eight areas.
Author
Julia Melo
Co-Founder
Julia is a co-founder of Meilynx, where she's building governance and cost observability for AI workloads — helping organizations understand, control, and scale AI adoption responsibly. She brings more than 16 years of engineering leadership, most recently as Director of Engineering at SeekOut, where she led Platform and AI-enabled product teams. Her earlier career spans finance-analytics leadership at BlackRock, HP, and a range of startups.
Catching AI Agent Drift Without Exposing Your Prompts
AI agents can change after approval without a new release. How permission and prompt drift happen, and how to catch them while prompts stay in your environment.
How a Proxy Governs MCP Agents
What runtime governance requires at the MCP layer, and how a proxy between agents and their tools provides it.
Does Generative AI Client Outreach Trigger SEC and FINRA Recordkeeping Rules?
Yes. Supervision, recordkeeping, and communications obligations attach the moment an AI-drafted message reaches a client. Here is what that means in practice.
The Questionnaire Banks Already Have, and AI Vendors Are About to Meet
A decade-old bank questionnaire most large tech vendors already answer every year now has AI governance built in, and most AI companies haven't seen it coming
The Rule That Isn't There
SR 26-2 rewrote the bank model-risk rulebook for the first time in 15 years, and left out the exact AI systems banks are racing to deploy.
Latest Breach Report Exposes an Enterprise AI Control Gap
AI-related breaches rose 61%, and financial services now carries the second-highest breach cost. The gap is in the workflow around AI, not the model.
Building Trust in Enterprise AI
Why governance and evidence are becoming prerequisites for adoption.
Hiring by Algorithm: The Compliance Landscape for AI in the Workplace
AI now screens resumes and recommends candidates for interview. States are building a legal framework to protect candidates and prevent discrimination.
Your Next Exam Will Ask About AI: Five Document Requests to Prepare For
AI is entering exam document requests. The five artifacts examiners will ask for, why screenshots fail as evidence, and how to prepare your AI audit trail.
Measuring AI ROI: The Unmeasured Middle Between Spend and Outcome
Fewer than 1% of companies report real AI ROI. They track spend and outcome but never the adoption and proficiency in between. Here's how to close the gap.
The Impact of AI on Data Breaches
The rising cost of shadow AI, weak access controls, and absent governance — and where enforcement actually closes the gap.