How AI Is Changing Institutional Client Workflows
Updated: 14 hours ago
As institutional workflows evolve, fintech and data providers need to understand how clients operate before deciding where their capabilities fit.
By Danielle Moore Jarnot, Founder, Moore Consulting
Former buy-side trader and technical analyst with experience across institutional trading, market structure, sales, strategy and enablement.
Key takeaways
Institutional client workflows are becoming more varied. MCP, APIs, AI tools and traditional platforms increasingly coexist inside the same firm.
That makes assumptions about client needs riskier. Segment, size and stated technology requirements do not tell you how the team actually works.
Client-facing teams need to understand the workflow and the business case. Is the client trying to reduce cost, consolidate workflow or create new revenue-generating capability?
AI is changing how institutional clients consume and use financial data while institutional workflows continue to vary widely across firms and teams.
A fundamental investment team may still work primarily across vendor applications, research documents, internal models and Excel. Another team may access much of the same data through APIs and internal applications. A third may be introducing AI assistants, Model Context Protocol (MCP) or agents into selected parts of an otherwise familiar process. Increasingly, those environments coexist inside the same institution.
For fintech, market data, analytics and infrastructure providers, the expanding range of delivery methods and AI capabilities makes solution fit more dependent on how the client actually operates. Client-facing teams need to understand how the work gets done today, what the client is trying to change and what business outcome is driving the change.
The same data provider can now sit inside several different workflows at the same client. Understanding which one you are solving for changes the commercial conversation.
AI is expanding financial data consumption across institutional workflows
MCP provides one of the clearest measures of how quickly institutional data consumption is changing. The 2026 Global Market Data Study from SIX and Coalition Greenwich found that 58% of 73 global buy-side and sell-side firms were engaged with or considering MCP as a way to make trusted market data accessible to AI-enabled workflows.
Recent S&P Global commentary provides a live measure of that shift. On September 14, CFO Eric Aboaf said the company was up to 500 MCP connectors, 50% more than the prior quarter. API and LLM call volume had increased approximately fivefold from Q1 to Q2 after another fivefold increase from Q4 to Q1. Aboaf also said clients are continuing to renew S&P’s proprietary platforms while asking for more data through newer channels.
LSEG’s H1 disclosures show the same multi-channel pattern. More than 200 customers were engaged with LSEG Everywhere by the end of the first half of 2026, accessing AI-ready data through MCP, multi-cloud environments or direct integration with their own AI stacks. During the same period, 17,000 users were active on AI Search inside Workspace.
AI-enabled data consumption is scaling across multiple channels
58%
Buy- and sell-side firms engaged with or considering MCP
Source: SIX / Coalition Greenwich
500
S&P Global MCP connectors as of September 14, up 50% QoQ
Source: S&P Global
200+
LSEG customers engaged with AI-ready distribution
Source: LSEG
17,000
Active users of AI Search inside LSEG Workspace
Source: LSEG
PitchBook adds an important behavioral signal. Roughly 20% of client accounts had accessed its premium MCP connector by mid-July. Morningstar has said that combined engagement across both the PitchBook platform and MCP gives it a better picture of how clients work because usage varies by workflow, from frequent access to episodic use around a transaction, fundraise or diligence exercise.
The commercial implication is bigger than MCP adoption. Institutional clients are adding new ways to consume data while continuing to use established applications, feeds and workflows. For a client-facing team, the question is which combination matters to this client and this use case.
The same AI ambition can sit on very different operating environments
AI adoption alone is a weak proxy for workflow maturity.
MSCI's September 2026 analysis of 130 private-market general partners found that 84% had progressed beyond the AI exploration stage. Yet only 12% had achieved centralized data flow. Forty-two percent still operated on partially connected technology stacks, 32% had standardized but largely manual environments and 13% remained entirely siloed.
AI adoption is outrunning infrastructure maturity
Private-market operating environment | Share of surveyed GPs |
Partially connected technology stack | 42% |
Standardized but largely manual | 32% |
Entirely siloed | 13% |
Centralized data flow | 12% |
Source: 2026 MSCI General Partner Survey, N=130.
The gap between AI adoption and workflow maturity can also exist within one large institution. A central technology or data organization may be building sophisticated AI infrastructure while individual investment, research or operating teams continue to rely on spreadsheets, established applications and manual processes.
Bank of America illustrates the distance that can exist between AI ambition and full implementation inside one institution. In its July earnings discussion, the bank reported more than 300 approved AI and machine-learning use cases, including 114 generative AI use cases, with 34 fully implemented. In his September 14 remarks, CEO Brian Moynihan also emphasized human ownership of AI-assisted work and controls around autonomous agents.
The spread between approved use cases and fully deployed applications illustrates how broad the phrase “using AI” has become. Approval, deployment, workflow integration and autonomy represent different stages of adoption, even within one institution.
PNC shows how ownership strategy can change the provider opportunity. At the Morgan Stanley U.S. Financials Conference in June, CEO Bill Demchak described PNC building its own “AI factory,” including GPU compute and internal language models, in part to reduce reliance on externally consumed tokens.
One institution may want a provider to supply trusted proprietary data into an environment it builds and controls. Another may want a more complete external solution. A third may require AI capability to fit inside an existing process with substantial human review.
For providers, the operating model behind an “AI-forward” client matters more than the label.
Client segment still leaves the workflow unanswered
Segmentation, personas and account characteristics remain valuable because they help identify likely problems, stakeholders and buying patterns. They become less useful when the commercial team treats them as evidence of how a particular client operates.
An API request establishes a delivery requirement. The commercial questions begin with what happens to the data after it arrives and how it enters the client’s workflow. MCP signals an AI-enabled consumption use case, which still needs to be connected to the users, systems and business problem involved. An agentic initiative adds another question: which parts of the capability does the institution intend to build and control itself?
The risk of misreading the client increases as the institutional buying group expands. In The Institutional Data Buyer Is Changing, we looked at the stakeholder side of that shift. As technology, data, governance and business functions become more involved, assumptions made early in an opportunity have more places to break later.
For client-facing teams, the work is to get underneath the technology request.
How does information move from source to user to analysis to decision today?
Where is the friction?
What is changing?
Which parts of the workflow will the institution own, and which does it expect a provider to solve?
A client's technology stack tells you what may be possible. Understanding the workflow tells you where your solution can create value.
Follow the workflow into the business case
Understanding the workflow still leaves one important question: why is the client willing to change it?
MSCI CFO Andrew Wiechmann gave a clear commercial framing of this issue in his September 14 remarks. He described two ways MSCI is expanding with asset managers under pressure: helping clients improve efficiency by replacing internal systems, displacing providers and consolidating spend, while also moving deeper into front-office workflows that support investment decisions and growth.
Are you selling cost reduction, workflow consolidation or revenue-generation capability?
Cost reduction may come from eliminating duplicated vendors, replacing internally maintained systems, reducing manual work or lowering the operating cost of an existing process.
Workflow consolidation may reduce handoffs, connect fragmented information, simplify the number of systems users move between or make data available inside an environment where the client already works.
Revenue-generation capability may help an investment team expand coverage, improve decision support, create new analytical capabilities, serve additional clients or build differentiated investment products and strategies.
The same data or technology may contribute to all three. The strongest buying argument will vary by client, and the stakeholder accountable for the outcome may change with it.
More capability raises the bar for client understanding
Financial data and fintech providers can now support institutional clients through established platforms and feeds alongside APIs, cloud delivery, MCP, AI search and emerging agentic capabilities. More ways to solve the problem also create more ways to recommend a technically valid solution that does not fit the client’s workflow. Client-facing teams therefore need to understand which capability fits, what the institution intends to own and what business outcome the solution is expected to support.
Recent management commentary shows how differently firms are responding to the same shift. S&P Global is growing proprietary platforms and AI-enabled distribution in parallel. MSCI is moving deeper into front-office workflows as clients pursue efficiency and growth. Bank of America is scaling AI while retaining explicit human accountability. PNC has chosen to own more of its AI infrastructure and development internally.
For client-facing teams, the requirement is broader than product knowledge. They need enough workflow fluency to qualify the problem, recognize where operating models are changing and determine what role the provider should play before product, technical and commercial resources are committed.
Five things client-facing teams need to establish
Current workflow: How does information move from source to user to analysis to decision today?
Friction: Where does the client lose time, confidence, coverage, capacity or efficiency?
Intended change: What is the client trying to improve, automate or enable?
Direction and ownership: What will remain central, what will change, and what does the client intend to build internally versus source externally?
Economic outcome: Is the business case cost reduction, workflow consolidation or revenue generation?
Better answers to those five questions produce a better-qualified client problem. As institutional workflows become more varied, that distinction will matter more than knowing which technology is generating the most attention.
Sources and methodology
This analysis draws on public 2026 research and disclosures from SIX/Coalition Greenwich, LSEG, Morningstar/PitchBook, MSCI and Bank of America, together with management commentary from S&P Global, MSCI and Bank of America at the Barclays 24th Annual Global Financial Services Conference on September 14, 2026. PNC commentary on internally owned AI infrastructure is from the Morgan Stanley U.S. Financials Conference in June 2026.
Survey populations and methodologies differ, so individual adoption figures should not be interpreted as directly comparable measures of the institutional market. Together, the evidence demonstrates rapid growth in AI-enabled data access alongside continued use of proprietary applications, APIs, manual workflows and institution-specific operating models.
Moore Consulting advises fintech, data, and infrastructure companies selling into financial services on GTM strategy, positioning and sales execution.
Moore Insights examines how revenue teams translate strategy into execution as complexity scales.
Comments