Microsoft has quietly expanded the capabilities of its Microsoft 365 Copilot, equipping licensed users with a suite of lesser-known but powerful tools—Mode Switcher, Copilot Pages, Researcher, Copilot Notebooks, and Prompt Gallery—that provide granular control over how the AI assistant responds and collaborates. While headlines often focus on Copilot’s integration with Word, Excel, and Teams, these underlying features give knowledge workers fine-tuned command over AI-driven research, content creation, and prompt reuse, making them indispensable for power users.

These tools are not new in the strictest sense; many have been available in preview or as part of Copilot for Microsoft 365 since its early rollouts. However, their placement and discoverability have improved, and Microsoft is now drawing attention to them as part of a broader push to help organizations extract more value from their Copilot subscriptions. For IT administrators, the enhanced toolset also brings fresh governance considerations, as each feature can influence data handling, collaboration patterns, and user behavior.

The Role of Lesser-Used Copilot Tools

Copilot for Microsoft 365 is most visibly experienced through its embedded assistive panes in apps like Word, where it drafts documents, or in Outlook, where it summarizes email threads. But beneath these surface-level experiences, the platform includes modular capabilities that function across the Microsoft Graph, the web, and internal data repositories. Mode Switcher, Copilot Pages, Researcher, Copilot Notebooks, and the Prompt Gallery belong to this cross-app layer. They are accessible from the Copilot chat interface in Microsoft Teams, Microsoft Edge, or the dedicated Copilot app experience.

Each tool addresses a distinct pain point in AI-assisted work: inconsistency in response style, lack of persistent collaborative canvases, shallow research outputs, the need to iterate on complex prompts, and the difficulty of reusing effective prompts across teams. By surfacing these tools more prominently, Microsoft is acknowledging that enterprise users require more than just inline text generation—they need flexible, multimodal AI interactions that adapt to diverse workflows.

Mode Switcher: Tailoring AI Responses

The Mode Switcher lets users toggle between different response styles: Precise, Balanced, and Creative. This feature, familiar from Bing Chat and the standard Copilot experience, is now more deeply integrated into Microsoft 365 Copilot. When working on a legal contract in Word, a user might switch to Precise mode to get concise, technically accurate suggestions. Later, while brainstorming marketing copy in a Copilot Page, they could flip to Creative mode for more imaginative output.

From an IT governance standpoint, the Mode Switcher presents both opportunity and risk. On one hand, it allows employees to calibrate the AI to their task, reducing the chance of off-brand messaging. On the other, Creative mode may generate content that feels less authoritative, requiring clear internal guidelines on when each mode is appropriate. Microsoft has not yet added admin-level controls to lock certain modes, leaving it to organizations to establish best practices through training and policy.

Copilot Pages: Collaborative AI Workspaces

Copilot Pages is an interactive, persistent canvas where users and Copilot co-create content in real time. Think of it as a blend of a whiteboard, a document, and a chat session. A user can ask Copilot to research a topic, and the AI will generate text, tables, and reference links directly onto the Page. Multiple team members can then edit, comment, and add their own prompts, turning the Page into a living document that evolves alongside the project.

This feature shines in scenarios like project planning, where inputs come from various stakeholders. A project manager might seed a Page with initial requirements, then invite engineers and designers to add constraints or ideas. Copilot can reconcile conflicting inputs, suggest compromises, and maintain a version history. Pages are stored in SharePoint, which means they inherit existing permissions, compliance policies, and eDiscovery capabilities—crucial for regulated industries.

The introduction of Copilot Pages also signals Microsoft’s ambition to redefine collaborative authoring. Instead of emailing drafts back and forth, teams can work in a shared space where the AI serves as both scribe and mediator. However, adoption may hinge on users’ willingness to shift from familiar tools like Word and Excel to this new paradigm.

Researcher: Deep-Dive Assistance

Researcher mode transforms Copilot into a thorough investigator, combining internal data from the Microsoft Graph with public web information. Unlike a standard chat query, which might return a few paragraphs, Researcher produces multi-source reports complete with citations. It’s especially useful for competitive analysis, market research, or preparing for executive briefings.

When a user invokes Researcher, Copilot explicitly states its reasoning steps, showing which sources it consulted and why. This transparency is critical for enterprise trust, as it allows users to verify facts before they make it into a final document. Researcher also respects the organization’s data boundaries—internal files are never mixed with public web results unless the user explicitly permits it.

For IT departments, Researcher raises questions about data egress and acceptable use. Even though the feature is designed to support grounded research, it can be misused for unauthorised queries or accessing rival information in ways that skirt company policy. Auditing capabilities in the Microsoft Purview compliance portal can help admins monitor Researcher usage, but proactive education remains the first line of defense.

Copilot Notebooks: Iterative Prompt Engineering

Copilot Notebooks offer a more technical interface for power users who need to refine complex, multi-step prompts. Unlike the single-shot chat experience, Notebooks allow users to build a sequence of prompts and responses, testing variations and observing how small changes affect the output. This is essentially a prompt engineering playground embedded within the Microsoft 365 ecosystem.

Data scientists, analysts, and developers stand to benefit most. For example, an analyst could use Notebooks to craft a prompt that extracts trends from sales data, iterating until the AI reliably produces the desired chart format and commentary. Once perfected, the prompt sequence can be saved and shared—bridging to the Prompt Gallery.

Notebooks also support parameterization, where users define variables that coworkers can fill in later. This makes the tool valuable for building reusable templates for common tasks like monthly reporting or customer insights summaries. However, the feature is currently less discoverable than others, tucked behind the Copilot chat interface, which may limit its uptake among less technical staff.

The Prompt Gallery addresses one of the biggest friction points in enterprise AI adoption: prompt inconsistency. Users often discover through trial and error that a specific wording yields far better results, but without a shared repository, that knowledge remains siloed. The Prompt Gallery lets individuals save effective prompts—along with notes on their intended use—and share them with colleagues or entire teams.

A marketing department, for instance, could create a curated gallery of prompts for social media posts, ad copy, and press releases. New hires can immediately adopt these tested prompts, maintaining brand voice consistency. Advanced users can also rate and comment on prompts, fostering a community-driven approach to improving AI interactions.

From a governance perspective, the Prompt Gallery is a double-edged sword. It promotes efficiency and consistency but could also spread biased or non-compliant prompts if not monitored. Organizations can nominate prompt curators and set up review workflows to vett gallery submissions. Microsoft is exploring features that would allow admins to designate sanctioned prompts and restrict sharing outside approved groups.

Impact on IT Governance and Productivity

Collectively, these tools shift AI usage from ad-hoc queries to structured, collaborative, and repeatable workflows. This evolution places new demands on IT and compliance teams. Key considerations include:

  • Data Security: Copilot Pages and Researcher interact with sensitive internal data. Organizations must ensure that SharePoint permissions and sensitivity labels are correctly configured, as Copilot respects these controls. Misconfigured permissions could inadvertently expose confidential information to end users.
  • Auditing: The Microsoft Purview compliance portal provides logs of Copilot interactions, but the richness of these logs varies by feature. For example, auditing a Researcher session yields more granular data than a simple chat query. IT teams need to understand what is logged and for how long.
  • User Training: The effectiveness of tools like Mode Switcher and Prompt Gallery hinges on user understanding. Without training, employees may default to the wrong mode or share poorly constructed prompts. Organizations should invest in workshops and accessible reference materials.
  • License Management: All these features are included in the Microsoft 365 Copilot license, but usage patterns can affect return on investment. IT can track adoption metrics via the Microsoft 365 admin center to see which features are being used and whether further promotion is needed.

Balancing empowerment with control is the central challenge. Overly restrictive policies will stifle the very productivity gains these tools promise. Too lenient an approach invites data leaks and off-brand outputs. A thoughtful, iterative governance model—where policies evolve based on real usage data and employee feedback—is the recommended path.

What This Means for Microsoft 365 Users

For everyday users, the immediate benefit is greater control over how Copilot behaves. The Mode Switcher makes AI responses more predictable. Copilot Pages reduces the friction of collaborative content creation. Researcher and Notebooks cater to more demanding tasks that require depth and precision. The Prompt Gallery democratizes best practices, ensuring that institutional knowledge about effective prompting spreads quickly.

These features also lower the barrier to entry for power-user scenarios that were previously the domain of advanced prompt engineers. A sales manager with no coding background can use Notebooks to fine-tune a quarterly review prompt, save it to the Gallery, and share it with their team—all without leaving the Microsoft 365 environment.

However, the onus is on organizations to communicate these capabilities clearly. Many licensed users remain unaware that tools like Pages and Researcher exist, often because they are buried in the Copilot chat interface or not actively promoted during onboarding. Microsoft could improve discoverability through in-app guidance and contextual prompts that appear when a user’s task aligns with a specific tool.

Looking ahead, the integration of these tools suggests a roadmap where Copilot becomes less of an assistant and more of a platform. Microsoft is likely to open APIs that let third-party developers embed their own prompts, skills, and connectors into the Gallery and Pages, creating an ecosystem of Copilot extensions—similar to how Teams apps extend collaboration. For now, the current feature set provides a solid foundation that rewards curious users and forward-thinking IT leaders alike.