Microsoft Elevates Excel: New Generative AI ‘Skills’ and Planning Modes Target the Finance Sector

In a significant move to solidify its dominance in enterprise productivity, Microsoft has unveiled a series of high-level updates for its 365 Copilot assistant specifically tailored for Microsoft Excel. These enhancements, which bridge the gap between simple automation and sophisticated financial modeling, are designed to transform Excel from a static spreadsheet tool into an agentic, data-driven analytical engine. By introducing customizable “skills” and a new “plan” mode, Microsoft is aiming to mitigate the “black box” nature of generative AI, providing finance professionals with the control and transparency required for high-stakes decision-making.

Main Facts: A New Era for Spreadsheet Automation

The latest suite of updates represents a shift in how Copilot interacts with complex datasets. Rather than functioning solely as a conversational chatbot that answers simple queries, Copilot is evolving into a task-oriented agent.

The headline features include:

• Customizable Skills: Users can now define, store, and deploy repeatable processes. Whether it is performing a complex discounted cash flow (DCF) analysis, conducting a variance analysis, or refreshing monthly reporting models, these skills allow users to codify institutional knowledge into a format Copilot can execute repeatedly.
• Plan Mode: To address the growing concern regarding AI hallucinations or unintended data manipulation, the new “Plan” mode forces Copilot to present its intended methodology before execution. It acts as a digital “reviewer,” asking the user to approve, edit, or clarify specific steps.
• Third-Party Data Connectivity: Recognizing that financial data rarely lives in a silo, Microsoft has enabled Copilot to pull real-time insights from third-party platforms, including Moody’s, CB Insights, Morningstar, and PitchBook.
• Auditability: Once actions are performed, Copilot links all changes within the chat interface, and these edits appear in Excel’s “Show Changes” pane, ensuring that AI-led actions are just as trackable as those performed by human analysts.

Chronology: The Evolution of Copilot in Excel

Microsoft’s journey to bring generative AI into the spreadsheet environment has been deliberate and iterative.

• Late 2024: Microsoft 365 Copilot reached general availability in Excel. This initial rollout allowed users to generate formulas, summarize data, and create visualizations through natural language queries.
• Early 2025: Microsoft began integrating “agentic” tools—capabilities that allow Copilot to perform multi-step operations independently. During this period, the company also introduced Python support within Excel, allowing users to perform sophisticated statistical analysis and machine learning directly within the grid.
• Mid-2025: The introduction of the “Copilot function” within cells allowed for more granular interactions, moving away from a side-panel-only experience.
• June 2026: Microsoft announced the current suite of “Finance Skills” and the “Plan” mode. This rollout marks the transition from general-purpose assistance to industry-specific professional workflows.

Supporting Data: Why Finance Professionals Need These Tools

The demand for these tools is driven by the sheer volume and complexity of data management in modern finance. According to Microsoft, finance professionals spend a disproportionate amount of time on “data hygiene”—formatting, reconciling, and structural modeling—rather than high-value strategic analysis.

The new “SKILL.md” framework is designed to address this. By allowing organizations to save standardized processes in OneDrive, Microsoft is essentially creating a library of “AI-ready” best practices. This ensures that when a junior analyst asks Copilot to perform a variance analysis, the AI adheres to the specific company template, formatting, and risk-adjustment parameters that the CFO’s office has mandated.

Furthermore, the integration of third-party data is a critical step in reducing the time analysts spend switching between browser tabs and Excel. By piping data directly from platforms like PitchBook or Morningstar, the workflow becomes fluid. A user can now command Copilot to “Build a competitive analysis table for [Company X] using the latest valuation data from PitchBook,” and the assistant handles the data ingestion, formatting, and initial calculation.

Official Responses: Insights from Microsoft Leadership

Brian Jones, Vice President for Excel at Microsoft, emphasized that the goal is not to replace the analyst, but to provide a “force multiplier” for their expertise. In a recent blog post, Jones noted, “Instead of starting from scratch each time, a skill guides Copilot through the steps, applying the right structure and formatting, and helping produce an output that is easier to review, reuse, and trust.”

This sentiment addresses the primary friction point for AI in finance: trust. By implementing the “Plan” mode, Microsoft is shifting the power dynamic. The AI is no longer the final decision-maker; it is a consultant that proposes a course of action. If Copilot decides to change a specific formula in a complex model, it must now present that plan to the user. This allows the human professional to audit the logic before the computer executes the calculation, maintaining the necessary human-in-the-loop oversight required for regulatory and internal compliance.

Implications: The Future of Financial Modeling

The implications of these updates for the financial services industry are profound.

1. Standardization vs. Flexibility

One of the greatest challenges in large financial institutions is the proliferation of inconsistent spreadsheets. When every analyst builds a DCF model differently, auditing becomes a nightmare. By utilizing standardized SKILL.md files, companies can enforce uniformity. This allows for a “gold standard” of modeling that is accessible to all, effectively democratizing high-level financial analysis across a firm.

2. The Rise of the "AI-Augmented Analyst"

The role of the junior analyst is set to change dramatically. Tasks that were once considered the “rite of passage”—the manual heavy lifting of data entry and standard reporting—are being automated. This forces a shift in skill sets: the value of a finance professional will increasingly depend on their ability to design, curate, and verify AI-driven models, rather than their speed in manual data manipulation.

3. Integration with the Ecosystem

The partnership with major financial data vendors signals that Microsoft is positioning Excel as the primary terminal for the modern financial professional. By connecting to data sources like Moody’s and CB Insights, Excel is directly competing with proprietary, closed-loop financial terminals. If a user can access institutional-grade data and analyze it with a customized AI agent inside the familiar Excel environment, the value proposition of external, expensive platforms may be tested.

4. Economic Barriers and Accessibility

Microsoft has maintained a two-tiered pricing structure to ensure scalability. Larger enterprises can access these features for $30 per user/month, while the Microsoft 365 Copilot Business plan—priced at $21 per user/month—targets smaller organizations with fewer than 300 employees. This aggressive pricing model suggests that Microsoft intends to make these tools a baseline requirement for the workforce, rather than a luxury for the elite.

Conclusion: A Paradigm Shift in Productivity

As Microsoft rolls out these features progressively, the landscape of financial modeling is poised for a significant transformation. The transition from “chat-based AI” to “process-based AI” is the most important development in the evolution of generative tools to date.

For the finance professional, the promise is clear: less time spent on the mechanics of spreadsheet management and more time spent on the nuance of financial strategy. However, the success of these tools will ultimately rest on the quality of the skills created and the vigilance of the human users who supervise the “Plan” mode. As we move into this new era of “frontier finance,” the synergy between human judgment and machine-speed processing will define the competitive advantage of the next decade.

While the technology is currently in its early stages of widespread adoption, the path forward is unmistakable. Microsoft is not merely updating a software application; it is rewriting the operating manual for the global financial sector. As these tools become more deeply embedded in daily workflows, the distinction between a “spreadsheet user” and an “AI-orchestrator” will likely disappear, leaving behind a new generation of professionals who are capable of doing more, with more accuracy, in a fraction of the time.