With trillions of dollars in client assets under management, Morgan Stanley Wealth Management serves millions of individuals and institutions globally. Known for blending high-touch advisory with cutting-edge technology, the organization sits at the intersection of tradition and transformation.

The first part of this three-part series delved into the practical execution of Morgan Stanley’s “One Firm” strategy, highlighting how connected data, enterprise intelligence, and a client-centric approach are powering the firm’s transformation and redefining leadership standards in the era of AI and advanced analytics.

In this second installment, Atul Dalmia, Chief Analytics Officer, Morgan Stanley Wealth Management, speaks with Acceldata CEO Rohit Choudhary about how AI is being deployed across Morgan Stanley, the tools that are proving most effective, and the deeply intentional framework the firm has established to govern innovation.

Meeting clients where they are

Morgan Stanley’s wealth management business spans three segments:

• Advice-driven (via financial advisors)
• Self-directed (such as E*TRADE, acquired in 2020)
• Workplace (managing stock plans and retirement solutions)

Each caters to clients with varying levels of financial experience, autonomy, and complexity. According to Dalmia, AI plays a crucial role in bridging these channels. “We use a propensity model to identify self-directed or workplace clients who feel they are ready to move to advice by using historical data,” he explains. These models draw on over 400 attributes to detect when a self-directed or workplace client may be ready for a financial advisor relationship.

AI does not just drive when to engage but also how, says Dalmia. “We target them using our relationship manager as well as additional interactions to make sure we have the conversation at the right time,” he adds. This allows Morgan Stanley to scale trust-based conversations with precision and personalization.

Productivity at the core

When it comes to enterprise AI, Dalmia believes success begins with practical utility. He highlights a suite of tools Morgan Stanley has launched to boost financial advisor productivity:

• AI-powered chatbot: Designed to provide advisors with “rapid access to all the Morgan Stanley intellectual capital,” this generative AI (GenAI) tool has been deeply integrated into the firm’s workflow. “98% of our financial advisors actively use this tool, and it has increased productivity,” says Dalmia.
• Debrief: A GenAI tool that summarizes client meetings — turning advisor notes into succinct emails and actionable next steps. Launched in mid-2024, it is now being rolled out to mobile and Salesforce platforms. “It has driven efficiency. It’s more like an eight or nine on a scale of ten in terms of impact,” Dalmia adds.

The company stresses ensuring that it controls what the AI works on. Dalmia states that most of the material used comes from internal sources, even if that means sacrificing some external accuracy.

He adds that the focus is on linking and labeling content so that, whenever needed, the origin of the information can be traced. From a productivity standpoint, this approach has been very effective, Dalmia states, as the analyses consistently show a clear increase in efficiency.

Bias, accuracy, and human-in-the-loop

With AI tools now augmenting core advisory services, he acknowledges the sensitive balance between automation and human expertise. “Right now we still view this as something that aids our financial advisors,” Dalmia says. Even tools like Debrief require advisor validation before sending outputs to clients. “We’re collecting feedback from them as to how useful or accurate the information has been.”

When asked whether Morgan Stanley envisions a future where humans are fully removed from the loop, Dalmia is cautious but optimistic. “It’s an evolving journey,” he says. The current approach ensures that AI supports but doesn’t replace human judgment.

With AI-driven personalization becoming a default expectation, how does a legacy brand like Morgan Stanley ensure its identity remains intact?

Dalmia believes personalization strengthens, not weakens, the firm’s brand. He explains that the organization looks at each client as an individual. “Even financial advisors treat clients individually.”

Predictive models already evaluate each client’s needs, and the firm offers a diversified product suite to support everyone — from emerging investors to ultra-high-net-worth individuals.

“We don’t look at it as brand dilution,” Dalmia affirms. Instead, it is a reaffirmation of the firm’s core ethos: meet every client where they are and provide the best solution for their unique journey.

AI governance at Morgan Stanley

In a landscape of fast-evolving generative AI capabilities, governance becomes non-negotiable. Dalmia outlines a robust framework for AI development and deployment at Morgan Stanley:

• Internal test bed: New ideas are tested before going into production.
• Governance forums: Every potential use case undergoes thorough evaluation by governance forums and must be approved through a formal governance committee to ensure only the most appropriate applications move forward.
• Controlled content: The data used within the system is tightly controlled, consisting solely of internally published content and research that has been vetted before deployment.

Thanks to a vast repository of high-quality historical content, the organization can deliver substantial value to clients. The governance model ensures that all AI use cases are carefully reviewed, while the data feeding into these tools remains both accurate and trusted.

Concluding, Dalmia states that explainability is central to the firm’s definition of success. Considerable effort goes into understanding what inputs drive AI recommendations and clearly identifying their sources. This rigorous approach ensures responsible AI deployment and reinforces the integrity of client-facing interactions.

CDO Magazine appreciates Atul Dalmia for sharing his insights with our global community.