LSEG + Microsoft: Practical AI Agents for Banks, Built on Licensed Market Data
LSEG is partnering with Microsoft to let financial institutions build AI agents directly on top of LSEG's licensed data inside Microsoft Copilot Studio. The integration supports the Model Context Protocol (MCP), improving connectivity between bank systems, third-party apps and AI agents.
LSEG's AI Ready Content and taxonomies draw on decades of normalized datasets. For banks, that means lower integration friction, clearer lineage and a safer path to production.
What's new
- Direct connection: LSEG customers can pipe licensed datasets into agents built in Copilot Studio for research, risk, trading and client workflows.
- Interoperability via MCP: An open standard that links tools, data and AI agents across your stack, reducing vendor lock-in and duplicate integrations.
- Low-code build: Copilot Studio enables teams to create and deploy agents into Microsoft 365 Copilot and other apps with enterprise controls.
"LSEG's partnership with Microsoft is transforming access to data for financial professionals with secure, seamless connectivity through MCP," said David Schwimmer, LSEG Group CEO. Nick Parker, Microsoft's Chief Business Officer and President, Worldwide Sales and Solutions, added, "By combining LSEG's trusted market data with Microsoft's cloud and AI capabilities, we're empowering customers to unlock deeper insights, accelerate decision-making and streamline complex workflows."
Why this matters for finance leaders
- Research and origination: Natural-language queries over decades of fundamentals, estimates, ownership and events with controlled citations and lineage.
- Risk and controls: Agent-driven monitoring, exposure summaries and alerts tied to internal and external data, with audit trails.
- Trading and execution support: Real-time signals, scenario prompts and post-trade analysis, gated by entitlements and compliance policies.
- Client service: On-demand briefings and Q&A grounded in licensed datasets and your firm's notes and models.
How to act now
- Inventory entitlements: Map which LSEG datasets your teams can use (and where MCP can safely expose them to agents).
- Stand up an MCP gateway: Connect a sandbox agent to a small, high-value dataset; log prompts, outputs and citations.
- Ship a minimum viable agent: Target one workflow (e.g., earnings-prep brief, sector comp sheet, risk factor roll-up). Measure time saved and error rates.
- Operationalize controls: Enforce access by role, redact sensitive fields, store conversations, and configure human-in-the-loop for decisions.
- Plan for scale: Define SLAs, monitoring and a release cadence. Treat prompts, tools and data contracts as versioned artifacts.
Adjacent partnerships to watch
- LSEG x Databricks: Brings AI-ready financial data into Databricks with Agent Bricks for real-time investment analytics, risk and trading workflows.
- LSEG x Rogo: Integrates LSEG datasets with Rogo's AI models and agents for search, analysis and automated banking workflows.
Macro check: tariffs, inflation and pricing behavior
Philadelphia Fed President Anna Paulson said tariff-related price increases have been "somewhat smaller than anticipated" and are unlikely to leave a lasting imprint on inflation. She cited a labor market that's discouraged churn, retailers prioritizing market share over passing through costs and already-restrictive monetary policy.
She still expects some additional goods inflation in the next few quarters as current and newly announced tariffs flow through. The takeaway: price sensitivity remains high, but broad second-round effects look contained for now.
Implications for banks and capital markets
- Rates and duration: With persistent inflation less likely from tariffs alone, scenario sets should keep a modest disinflation path in view.
- Retailer margins: Expect selective, category-level price moves over blanket hikes; watch elasticity and traffic into the holiday quarter.
- Credit: Margin defense via pricing may be limited; focus on inventory turns, cash conversion cycles and working-capital strain.
- Risk models: Update goods-inflation assumptions and pass-through coefficients; monitor sectors with tariff-sensitive inputs.
Useful resources
- Microsoft Copilot Studio - build, customize and deploy enterprise agents.
- Model Context Protocol (MCP) - open standard for connecting tools and data to AI agents.
- AI tools for Finance - curated options to accelerate research, risk and reporting.
Bottom line
AI agents are moving from demos to desks. With LSEG's licensed data, MCP connectivity and Copilot Studio, banks can build targeted agents that respect entitlements, reduce manual prep and tighten controls. Start small, measure rigorously, and scale what proves useful.
Your membership also unlocks:
