Generative AI inside an advisor's governed workflow.
AI @ Morgan Stanley retrieves approved internal knowledge and turns consented client meetings into notes and draft follow-ups, while the advisor remains responsible for the final work.
Selection rationale
Why it was selected
Generative AI in finance needs a narrow job, controlled information, and a person who remains accountable. Morgan Stanley built its tools around approved firm knowledge, client consent, and advisor review rather than autonomous advice.
Morgan Stanley reports that more than 98% of advisor teams actively use the Assistant. The system expanded from 7,000 knowledge items to 100,000 documents, and the share of firm documents advisors could effectively search rose from 20% to 80%.Morgan Stanley Debrief launch, June 2024OpenAI case study
The panel favored production adoption, consent, controlled retrieval, and human review over raw automation volume. The adoption and access figures come from Morgan Stanley and its technology provider, not an independent audit.
Retrieves approved research and turns meetings into advisor-reviewed follow-up work.
Requires client consent for meeting capture and leaves final output with the advisor.
Reached active use across more than 98% of Morgan Stanley advisor teams.