This offering will be available through a purpose-built Model Context Protocol (MCP) application and will make MAS accessible in Claude Desktop, Claude.ai and Claude Enterprise, the companies said in a Thursday (April 9) press release.
Moody’s agents will support credit analysis for financial institutions as well as compliance workflows spanning entity profiling, ownership structure mapping, adverse media screening and sanctions checks. These will be rendered as interactive reports directly within Claude, according to the release.
Additional capabilities across risk monitoring and portfolio intelligence will follow, as a series of Moody’s agentic workflows are planned for the Claude environment.
Each agent is grounded in Moody’s connected intelligence, and all outputs are valid, explainable and auditable to meet the standard required in regulated environments, per the release.
“The institutions that will lead in an AI-driven world are those that build on intelligence that can be trusted, defended and acted upon,” Cristina Pieretti, head of digital content and innovation at Moody’s, said in the release. “Moody’s provides that intelligent layer — connected, decision-grade and now available directly in the Claude environment where our customers are already working.”
Advertisement: Scroll to Continue
The companies also announced in the release that Moody’s is deploying Claude Enterprise, Claude Code and Claude Desktop in its own operations to accelerate product development.
Kate Jensen, head of Americas at Anthropic, said in the release: “Claude is built for work where the stakes are high and the outputs need to be defensible, and that’s exactly what credit and compliance teams face every day. Moody’s is going all in by bringing Moody’s Agentic Solutions natively into Claude for their customers and deploying Claude in their own operations.”
It was reported in September 2025 that Moody’s Agentic Solutions agents reduced the preparation time of credit memos from 40 hours to two minutes. MAS breaks down credit memo preparation into specialized tasks handled by different agents that run in parallel, with one verifying entries, another extracting data from filings, and others assessing peer comparisons and risks.
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.
