Leonardo has begun a major workplace transformation program with Microsoft Italia that could reshape how roughly 50,000 employees collaborate, find information, manage documents, and use generative AI across one of Europe’s most strategically significant aerospace, defence, and security groups. The initiative, called “Digital Horizon – A New Era of Collaboration with Microsoft 365,” moves the company further toward cloud-based productivity and introduces Microsoft 365 Copilot as a contextual AI assistant within its corporate environment. Leonardo’s announcement frames the deployment as both a technology modernization effort and a cultural change program built around collaboration, training, intelligence, and security.
For Windows and Microsoft 365 administrators, the significance is not simply the scale of the rollout. It is the character of the organization making it. Leonardo operates across aerospace, defence, electronics, helicopters, cyber security, space, and other high-value technology domains, where document permissions, intellectual property boundaries, supply-chain coordination, and controlled communications are central operational concerns. The company describes itself as a global leader in aerospace, defence, and security, delivering interconnected technologies for people, systems, and infrastructure. Leonardo’s corporate overview makes clear that this is an environment in which a casual “turn on the AI assistant” approach would be inadequate.
Digital Horizon is therefore a useful case study in what enterprise AI adoption increasingly looks like: a broad collaboration-platform migration paired with phased enablement, internal champions, governance considerations, and practical instruction on when employees should use each tool.
Microsoft 365 has long been more than a collection of Office applications. In the enterprise, it is an interconnected productivity layer built around identity, messaging, meetings, files, collaboration spaces, compliance controls, and workflow services. That wider platform view is central to Leonardo’s stated plan.
The company says Digital Horizon will provide its people with modern work tools within Microsoft Teams, while using OneDrive and Outlook in the Microsoft cloud to synchronize documents and communications. The intended outcome is real-time collaboration and more seamless remote work across the Group. Leonardo’s press release explicitly identifies Teams, OneDrive, Outlook, Microsoft 365, and Copilot as the technological foundations of the program.
That matters because collaboration transformation is rarely about deploying one app. A Teams rollout that is disconnected from document governance becomes a messaging project. A OneDrive deployment without clear sharing practices can become a proliferation of files and links. An AI rollout that is detached from identity controls and information architecture can magnify problems that already existed quietly in the organization.
Digital Horizon appears designed to address those dependencies together. It groups the move to Microsoft 365 cloud productivity and the introduction of generative AI under a single program, rather than presenting Copilot as a standalone experiment. That is a more mature framing than the consumer-style narrative of AI as a magic writing button.
Microsoft describes Microsoft 365 Copilot as an AI-powered work tool that responds to prompts using real-time information, internet content where configured, and organizational content that the user is permitted to access. Its capabilities span familiar Microsoft 365 applications including Word, Excel, PowerPoint, Outlook, Teams, Loop, and OneNote. Microsoft’s Copilot overview explains that Copilot can use Microsoft Graph-based work context such as emails, chats, documents, and meetings, while respecting the individual user’s permissions.
In a well-governed deployment, that contextual awareness is the feature. In a poorly governed deployment, it is the risk.
The practical benefit is straightforward. Employees can coordinate around a project without forcing every decision into long email chains, while documents can be worked on in context rather than endlessly passed as detached attachments. This is particularly valuable in large organizations where work must cross engineering, procurement, HR, program management, administration, legal, and operational teams.
Yet Teams alone is not a solution to organizational fragmentation. Its success depends on disciplined workspace design: clear ownership, appropriate lifecycle controls, sensible channel structure, external collaboration rules, and a consistent decision about which records belong in chat, email, Teams channels, SharePoint sites, or formal business systems.
Leonardo’s “more connected and integrated” objective is promising precisely because it recognizes that tools should support a new way of working, not merely digitize older habits. As Antonio Liotti, Leonardo’s Chief People & Organization Officer, put it, the effort is intended to improve collaboration, information sharing, and innovation while reducing time spent on repetitive work. Leonardo’s announcement emphasizes that employees should be able to focus more of their time on work with greater added value.
Copilot’s usefulness depends on accessible, current, and appropriately organized work data. When the authoritative version of a document sits on an inaccessible local drive, in a fragmented network share, or in a mailbox folder known only to one employee, neither modern collaboration nor AI assistance can deliver its full value.
At the same time, moving information into cloud collaboration systems requires deliberate records management and data classification. Organizations need to decide:
The real value proposition is broader than document drafting. In Microsoft’s own description, Copilot can help create and summarize Word content, suggest formulas and insights in Excel, summarize email threads in Outlook, and assist with meetings and chats in Teams. Microsoft’s product documentation also notes that Copilot can retrieve work-context information from the Microsoft Graph, subject to the user’s existing permissions.
For a group the size of Leonardo, several early use cases are immediately credible:
The company plans short training modules, guidelines, support tools, advanced prompting education, and an introduction to AI agents. It will also use a “When to Use What” smart compass designed to guide employees through the available digital ecosystem. Leonardo’s announcement describes the training component as integral to helping staff navigate the new tools, rather than a peripheral afterthought.
That “when to use what” concept deserves attention. Modern knowledge workers often face a confusing overlap of tools:
A weak request might ask Copilot to “summarize this project.” A stronger request identifies the audience, purpose, desired length, source material, and constraints: summarize only the attached documents, flag open risks, distinguish confirmed decisions from proposed actions, and cite the relevant source passage where possible.
This is especially important in an industrial and security-sensitive environment. Users need habits that discourage vague, unreviewed AI output from being treated as authoritative. They also need practical awareness of when not to include sensitive information in a request, when to use approved internal sources only, and when the proper response is to escalate to a subject-matter expert.
Microsoft’s own Copilot deployment guidance recommends assessing organizational readiness, preparing data, supporting users with training resources, and gathering feedback after licenses are assigned. Microsoft’s adoption and onboarding guide treats enablement as a continuous operational responsibility, not as a single launch event.
This model aligns closely with Microsoft’s recommended use of early adopters and champions. Microsoft advises organizations to create early-adopter groups, observe how they find value in Copilot, and use their experience to help peers adopt the technology in context. Microsoft’s Copilot setup guidance notes that champions can provide a scalable, peer-oriented way to handle questions and develop expertise across business groups.
The distinction is important. A well-run champion network does not simply promote usage statistics. It translates the platform into role-specific practices:
Microsoft 365 Copilot inherits many of the security and compliance foundations already present in Microsoft 365. Microsoft states that Copilot accesses only the data users are authorized to access and respects existing organization-level controls for compliance, privacy, and data residency. Microsoft’s security documentation also says that Copilot can inherit user permissions, sensitivity labels, and retention policies.
That is an important protection—but it is not a guarantee that all underlying data is correctly governed.
Microsoft explicitly warns that overshared or poorly governed content can affect Copilot results and increase risk because Copilot works within the permissions and access controls that already exist. Microsoft’s Copilot security guidance is clear: inherited permissions protect against unauthorized access, but they can also expose content that had been shared too widely in the first place.
This is a crucial point for any organization following Leonardo’s path. Generative AI can make unstructured content easier to find, summarize, and connect. That is productive when permissions are correct. It is problematic when legacy sites, broad links, inactive project spaces, or vaguely defined groups expose material beyond the audience originally intended.
The right response is not to abandon AI. It is to treat the rollout as a catalyst for permission cleanup and data governance improvement.
Microsoft’s guidance for a secure Copilot foundation centers on three pillars:
For Leonardo, that approach should be especially relevant. The Group’s ambition to make collaboration more integrated must be balanced against the need to segment sensitive data by program, role, location, contractual requirement, and business purpose.
The company also now provides a Copilot security dashboard focused on data protection, oversharing, and compliance insights, alongside a broader AI security dashboard that can surface risk signals across Microsoft and third-party AI tools. Microsoft’s security guidance describes these tools as ways to monitor and act on AI-related risk across the enterprise.
A stronger measurement model would distinguish among four categories.
Those thresholds are Microsoft’s product analytics definitions, not universal business-performance benchmarks. Still, they illustrate the right principle: sustained, purposeful use matters more than launch-week excitement.
That approach has several clear strengths:
The strongest outcome will come if Digital Horizon remains disciplined about its stated “gradual integration” of AI. That means piloting high-value scenarios, learning from users, measuring impact, correcting data-access issues, and scaling only where controls and business benefits are demonstrably strong.
For IT teams, this changes the deployment conversation. The question is no longer simply whether to enable Copilot or standardize on Teams. The real questions are:
Leonardo’s decision to combine Microsoft 365 cloud collaboration, Copilot, training, and a Digital Pioneers network gives the program a credible foundation. Its lasting impact will depend on whether that foundation is maintained through rigorous information governance, role-specific enablement, meaningful outcome measurement, and a willingness to treat AI as an assistive capability that strengthens human expertise rather than substitutes for it.
That is the practical digital horizon now facing every large Microsoft 365 organization: not just adopting AI, but building the secure, collaborative, and accountable workplace in which AI can genuinely be trusted.
For Windows and Microsoft 365 administrators, the significance is not simply the scale of the rollout. It is the character of the organization making it. Leonardo operates across aerospace, defence, electronics, helicopters, cyber security, space, and other high-value technology domains, where document permissions, intellectual property boundaries, supply-chain coordination, and controlled communications are central operational concerns. The company describes itself as a global leader in aerospace, defence, and security, delivering interconnected technologies for people, systems, and infrastructure. Leonardo’s corporate overview makes clear that this is an environment in which a casual “turn on the AI assistant” approach would be inadequate.
Digital Horizon is therefore a useful case study in what enterprise AI adoption increasingly looks like: a broad collaboration-platform migration paired with phased enablement, internal champions, governance considerations, and practical instruction on when employees should use each tool.
Background: From Office Productivity to an Integrated Digital Workplace
Microsoft 365 has long been more than a collection of Office applications. In the enterprise, it is an interconnected productivity layer built around identity, messaging, meetings, files, collaboration spaces, compliance controls, and workflow services. That wider platform view is central to Leonardo’s stated plan.The company says Digital Horizon will provide its people with modern work tools within Microsoft Teams, while using OneDrive and Outlook in the Microsoft cloud to synchronize documents and communications. The intended outcome is real-time collaboration and more seamless remote work across the Group. Leonardo’s press release explicitly identifies Teams, OneDrive, Outlook, Microsoft 365, and Copilot as the technological foundations of the program.
That matters because collaboration transformation is rarely about deploying one app. A Teams rollout that is disconnected from document governance becomes a messaging project. A OneDrive deployment without clear sharing practices can become a proliferation of files and links. An AI rollout that is detached from identity controls and information architecture can magnify problems that already existed quietly in the organization.
Digital Horizon appears designed to address those dependencies together. It groups the move to Microsoft 365 cloud productivity and the introduction of generative AI under a single program, rather than presenting Copilot as a standalone experiment. That is a more mature framing than the consumer-style narrative of AI as a magic writing button.
Microsoft describes Microsoft 365 Copilot as an AI-powered work tool that responds to prompts using real-time information, internet content where configured, and organizational content that the user is permitted to access. Its capabilities span familiar Microsoft 365 applications including Word, Excel, PowerPoint, Outlook, Teams, Loop, and OneNote. Microsoft’s Copilot overview explains that Copilot can use Microsoft Graph-based work context such as emails, chats, documents, and meetings, while respecting the individual user’s permissions.
In a well-governed deployment, that contextual awareness is the feature. In a poorly governed deployment, it is the risk.
What Leonardo Is Actually Deploying
Microsoft Teams as the Collaboration Hub
Leonardo’s announcement places Microsoft Teams at the center of the employee experience. For an organization distributed across functions, locations, and national boundaries, Teams provides a shared interface for meetings, chat, channels, calls, document collaboration, and increasingly AI-assisted work.The practical benefit is straightforward. Employees can coordinate around a project without forcing every decision into long email chains, while documents can be worked on in context rather than endlessly passed as detached attachments. This is particularly valuable in large organizations where work must cross engineering, procurement, HR, program management, administration, legal, and operational teams.
Yet Teams alone is not a solution to organizational fragmentation. Its success depends on disciplined workspace design: clear ownership, appropriate lifecycle controls, sensible channel structure, external collaboration rules, and a consistent decision about which records belong in chat, email, Teams channels, SharePoint sites, or formal business systems.
Leonardo’s “more connected and integrated” objective is promising precisely because it recognizes that tools should support a new way of working, not merely digitize older habits. As Antonio Liotti, Leonardo’s Chief People & Organization Officer, put it, the effort is intended to improve collaboration, information sharing, and innovation while reducing time spent on repetitive work. Leonardo’s announcement emphasizes that employees should be able to focus more of their time on work with greater added value.
OneDrive and Outlook in the Microsoft Cloud
Digital Horizon also relies on the cloud synchronization of documents and communications through OneDrive and Outlook. This element may sound conventional beside generative AI, but it is arguably the essential infrastructure behind the project.Copilot’s usefulness depends on accessible, current, and appropriately organized work data. When the authoritative version of a document sits on an inaccessible local drive, in a fragmented network share, or in a mailbox folder known only to one employee, neither modern collaboration nor AI assistance can deliver its full value.
At the same time, moving information into cloud collaboration systems requires deliberate records management and data classification. Organizations need to decide:
- Which files belong in a personal OneDrive versus a shared SharePoint or Teams workspace.
- Who owns a workspace after a project ends or an employee changes roles.
- Which documents may be shared externally.
- How sensitivity labels, retention rules, and access reviews apply to technical, commercial, and personal information.
- When a Teams chat contains a business record that needs to be preserved elsewhere.
Microsoft 365 Copilot as the AI Layer
The high-profile aspect of the Leonardo-Microsoft collaboration is the planned use of Microsoft 365 Copilot. Leonardo says individuals and teams will have access to an AI assistant integrated into the corporate context, capable of supporting the creation, analysis, and synthesis of content in what the company calls a secure and trusted environment. Leonardo’s press release identifies Copilot as the mechanism for gradually integrating AI into work processes.The real value proposition is broader than document drafting. In Microsoft’s own description, Copilot can help create and summarize Word content, suggest formulas and insights in Excel, summarize email threads in Outlook, and assist with meetings and chats in Teams. Microsoft’s product documentation also notes that Copilot can retrieve work-context information from the Microsoft Graph, subject to the user’s existing permissions.
For a group the size of Leonardo, several early use cases are immediately credible:
- Meeting recap and action extraction from Teams meetings.
- Email-thread summarization for complex cross-functional discussions.
- Drafting and refining internal communications with clearer tone and structure.
- Document synthesis across approved project materials.
- Spreadsheet assistance for spotting patterns and preparing explanations of data.
- Search and knowledge discovery across content users are already authorized to see.
- Structured first drafts of agendas, plans, briefings, and recurring reports.
The Training Strategy Is One of the Program’s Strongest Features
Technology deployments often fail not because the platform is weak, but because employees are expected to infer new working practices for themselves. Leonardo’s announcement suggests a more deliberate approach.The company plans short training modules, guidelines, support tools, advanced prompting education, and an introduction to AI agents. It will also use a “When to Use What” smart compass designed to guide employees through the available digital ecosystem. Leonardo’s announcement describes the training component as integral to helping staff navigate the new tools, rather than a peripheral afterthought.
That “when to use what” concept deserves attention. Modern knowledge workers often face a confusing overlap of tools:
- Use Outlook for formal external and internal correspondence.
- Use Teams chat for rapid, direct coordination.
- Use Teams channels for visible, team-based project discussion.
- Use OneDrive for personal work files and individual sharing.
- Use shared workspaces for collaborative documents with team ownership.
- Use Copilot to accelerate discovery, drafting, analysis, and synthesis—but not to bypass verification or access controls.
Why Prompting Education Matters
Prompting is sometimes overstated as a specialist discipline. Employees do not need to become prompt engineers to get value from Copilot. They do, however, need to understand how to make requests that are specific, contextual, bounded, and reviewable.A weak request might ask Copilot to “summarize this project.” A stronger request identifies the audience, purpose, desired length, source material, and constraints: summarize only the attached documents, flag open risks, distinguish confirmed decisions from proposed actions, and cite the relevant source passage where possible.
This is especially important in an industrial and security-sensitive environment. Users need habits that discourage vague, unreviewed AI output from being treated as authoritative. They also need practical awareness of when not to include sensitive information in a request, when to use approved internal sources only, and when the proper response is to escalate to a subject-matter expert.
Microsoft’s own Copilot deployment guidance recommends assessing organizational readiness, preparing data, supporting users with training resources, and gathering feedback after licenses are assigned. Microsoft’s adoption and onboarding guide treats enablement as a continuous operational responsibility, not as a single launch event.
The Value of Digital Pioneers
Leonardo’s proposed network of “Digital Pioneers” may prove to be the most scalable part of the initiative. These colleagues are intended to support adoption and circulate best practices across the organization. Leonardo’s announcement positions them as a distributed support network rather than a centralized IT help desk substitute.This model aligns closely with Microsoft’s recommended use of early adopters and champions. Microsoft advises organizations to create early-adopter groups, observe how they find value in Copilot, and use their experience to help peers adopt the technology in context. Microsoft’s Copilot setup guidance notes that champions can provide a scalable, peer-oriented way to handle questions and develop expertise across business groups.
The distinction is important. A well-run champion network does not simply promote usage statistics. It translates the platform into role-specific practices:
- A project coordinator may learn how to generate a meeting-action digest.
- An HR professional may use Copilot to organize a policy draft while preserving review steps.
- A finance team may identify appropriate ways to analyze approved spreadsheet data.
- A manager may use Teams and Copilot to improve meeting preparation without creating an unmanageable archive of AI-generated content.
- A security-aware employee may learn to recognize when a request involves information that needs tighter handling.
Security Is the Critical Test, Not a Marketing Footnote
Leonardo has made intelligence and security central themes of Digital Horizon. That emphasis is essential. A company operating in aerospace, defence, and security cannot view collaboration technology as separate from information protection.Microsoft 365 Copilot inherits many of the security and compliance foundations already present in Microsoft 365. Microsoft states that Copilot accesses only the data users are authorized to access and respects existing organization-level controls for compliance, privacy, and data residency. Microsoft’s security documentation also says that Copilot can inherit user permissions, sensitivity labels, and retention policies.
That is an important protection—but it is not a guarantee that all underlying data is correctly governed.
The Oversharing Problem
The most important risk in an enterprise Copilot deployment is often not that the AI “breaks” a permission boundary. It is that the boundary was already too broad.Microsoft explicitly warns that overshared or poorly governed content can affect Copilot results and increase risk because Copilot works within the permissions and access controls that already exist. Microsoft’s Copilot security guidance is clear: inherited permissions protect against unauthorized access, but they can also expose content that had been shared too widely in the first place.
This is a crucial point for any organization following Leonardo’s path. Generative AI can make unstructured content easier to find, summarize, and connect. That is productive when permissions are correct. It is problematic when legacy sites, broad links, inactive project spaces, or vaguely defined groups expose material beyond the audience originally intended.
The right response is not to abandon AI. It is to treat the rollout as a catalyst for permission cleanup and data governance improvement.
Microsoft’s guidance for a secure Copilot foundation centers on three pillars:
- Remediate oversharing.
- Set up guardrails.
- Meet regulatory obligations.
For Leonardo, that approach should be especially relevant. The Group’s ambition to make collaboration more integrated must be balanced against the need to segment sensitive data by program, role, location, contractual requirement, and business purpose.
Controls That Need Continuous Attention
A secure Microsoft 365 Copilot deployment is an operating model, not a checkbox. The practical control set should include:- Strong identity protection, including multifactor authentication and conditional access.
- Least-privilege group membership and disciplined access reviews.
- Sensitivity labels that classify and protect documents based on their content.
- Data loss prevention policies for sensitive categories of information.
- Audit logging for user and administrator activity.
- External-sharing restrictions appropriate to the data and collaboration scenario.
- Lifecycle management for Teams, SharePoint sites, files, and inactive workspaces.
- Agent governance, particularly as employees begin creating or using task-specific AI agents.
- Human review standards for AI-generated work that informs decisions, technical documentation, or formal communications.
The company also now provides a Copilot security dashboard focused on data protection, oversharing, and compliance insights, alongside a broader AI security dashboard that can surface risk signals across Microsoft and third-party AI tools. Microsoft’s security guidance describes these tools as ways to monitor and act on AI-related risk across the enterprise.
Measuring Value Requires More Than Counting Licenses
Digital Horizon’s success should not be measured by how many Microsoft 365 accounts are provisioned or how many Copilot prompts employees submit. Those figures may indicate reach, but they do not prove business value, safer collaboration, or better employee experience.A stronger measurement model would distinguish among four categories.
1. Adoption
Basic adoption answers whether employees are actually using the platform:- Active Teams users.
- Active Copilot users.
- Adoption by function, region, and role.
- Usage by application and feature.
- Completion of training modules.
- Participation in Digital Pioneer-led sessions.
2. Habit Formation
Usage should be interpreted carefully. An employee who tries Copilot once is not necessarily benefiting from it. Microsoft’s reporting framework distinguishes between novice, habitual, and power users based on both frequency and consistency of Copilot activity. Microsoft’s adoption report guidance defines a power user, for example, as someone averaging at least 15 Copilot actions per week and using it in at least nine of the prior 12 weeks.Those thresholds are Microsoft’s product analytics definitions, not universal business-performance benchmarks. Still, they illustrate the right principle: sustained, purposeful use matters more than launch-week excitement.
3. Operational Outcomes
Leonardo should connect platform adoption to observable workflow improvements. Useful measures could include:- Reduced time spent preparing recurring meeting summaries.
- Faster retrieval of approved project information.
- Fewer duplicated files and email attachments.
- Shorter cycles for routine drafting and review.
- Better action tracking after meetings.
- Higher employee satisfaction with collaboration tools.
- Reduced reliance on unsupported or shadow AI services.
4. Security and Governance Outcomes
A mature scorecard should also report the controls that keep the collaboration environment trustworthy:- High-risk sharing links identified and remediated.
- Sensitive sites properly labeled and access-controlled.
- Audit coverage and policy compliance.
- Number and type of data-loss prevention events.
- AI agents approved, monitored, modified, or retired.
- Training completion for high-risk roles.
- Security incidents associated with collaboration content or AI use.
Why Digital Horizon Is a Significant Microsoft 365 Story
The Leonardo-Microsoft Italia partnership is notable because it treats digital transformation as a compound challenge. The Group is not merely adopting Teams, moving files to OneDrive, or giving employees an AI chatbot. It is bringing together cloud productivity, collaboration redesign, AI assistance, skills development, internal advocacy, and security governance.That approach has several clear strengths:
- Scale with purpose: Reaching approximately 50,000 people creates the possibility of consistent collaboration practices across a major international group. Leonardo’s announcement
- AI embedded in daily tools: Copilot’s integration with familiar Microsoft 365 applications lowers the friction of adoption compared with introducing a disconnected AI platform. Microsoft’s Copilot overview
- People-first enablement: The training modules, guidance, prompting education, and Digital Pioneers model acknowledge that work habits—not software availability—determine outcomes.
- Security-aware framing: Leonardo’s public emphasis on trusted, secure use is appropriate for the sectors in which it operates.
- Potential for continuous improvement: Microsoft’s reporting and deployment tools offer ways to observe adoption, investigate governance issues, and refine the program over time. Microsoft’s Copilot setup guidance
The strongest outcome will come if Digital Horizon remains disciplined about its stated “gradual integration” of AI. That means piloting high-value scenarios, learning from users, measuring impact, correcting data-access issues, and scaling only where controls and business benefits are demonstrably strong.
The Bigger Lesson for Enterprise Windows and Microsoft 365 Environments
Leonardo’s Digital Horizon program reflects a broader shift in enterprise computing. The modern workplace is no longer defined primarily by the Windows desktop, the Office suite, or email. It is defined by a connected work graph of identities, files, meetings, chats, business data, security policy, and now AI-assisted reasoning.For IT teams, this changes the deployment conversation. The question is no longer simply whether to enable Copilot or standardize on Teams. The real questions are:
- Is the data estate ready?
- Are permissions accurate?
- Are employees trained to use AI responsibly?
- Is there a clear policy for agents and automation?
- Can IT measure adoption without confusing activity with value?
- Can the organization detect and correct oversharing before it becomes an AI-enabled discovery problem?
Leonardo’s decision to combine Microsoft 365 cloud collaboration, Copilot, training, and a Digital Pioneers network gives the program a credible foundation. Its lasting impact will depend on whether that foundation is maintained through rigorous information governance, role-specific enablement, meaningful outcome measurement, and a willingness to treat AI as an assistive capability that strengthens human expertise rather than substitutes for it.
That is the practical digital horizon now facing every large Microsoft 365 organization: not just adopting AI, but building the secure, collaborative, and accountable workplace in which AI can genuinely be trusted.