Embracing AI and Modern Management: Key Insights from MMS 2025

May 17, 2025

Embracing AI and Modern Management: Key Insights from MMS 2025

Introduction: The New Normal for IT Management

Modern endpoint management has undergone a radical transformation by 2025. At this year's Midwest Management Summit (MMS 2025), a clear message emerged: cloud-native management and AI-driven automation are no longer futuristic concepts, but current best practices. Senior IT engineers and technical managers are rethinking management frameworks to integrate cloud services, intelligent copilots, and automated security into their IT strategies. This blog revisits our core IT management topics – cloud-first modern management, security, and automation – and updates them with the latest developments in AI and Copilot technologies as highlighted at MMS 2025. The goal is a forward-looking guide (300–500 level) to help technical leaders steer their teams through this new era of intuitive, AI-augmented IT management.

Modern Management 2.0: Cloud-First and Always Up-to-Date

Modern management refers to managing endpoints (desktops, mobiles, cloud PCs) via cloud services instead of traditional on-premises tools. Over the past few years, organizations have steadily shifted from Group Policy and Configuration Manager (ConfigMgr) toward cloud solutions like Microsoft Intune for policy, configuration, and software deployment. In 2025, this cloud-first approach is the norm rather than the exception. One major change with this model is the pace of updates and the loss of direct control over backend changes – cloud services update continuously without admin intervention. As one MMS 2025 session noted, “we are not in control of backend changes anymore” in an Intune-managed world. This requires IT teams to adjust their processes: instead of scheduling patch cycles or service packs, they focus on continuous monitoring and rapid response to changes that Microsoft rolls out automatically.

Intune and Windows 11 have matured hand-in-hand. Windows 11 is built with cloud integration in mind, and many new features light up only with MDM (mobile device management) policies in place. A practical consequence is that troubleshooting issues on modern managed clients looks different than in the old days. With ConfigMgr and on-prem Group Policy, IT had a trove of logs and complete visibility on server/client interactions. In contrast, troubleshooting Intune-managed devices often means parsing client-side logs (e.g. MDMDiagCollector logs, event viewer) and understanding SyncML MDM transactions – a new skill for many admins. MMS 2025 talks emphasized using cloud-based tools and data to diagnose problems. For example, admins now rely on Intune’s built-in troubleshooting pane, Graph API queries, and even AI assistance (more on that below) instead of digging through on-prem log files. Audit logs and service health messages from Microsoft are now essential references whenever “something isn’t working as expected,” since the service might have changed under the hood. The guidance for teams is clear: invest in upskilling on cloud management tooling, and build workflows that assume change is constant.

Another big theme in modern management is the concept of “co-management” evolving into full cloud management. Early on, many enterprises kept a foot in both worlds (using both ConfigMgr and Intune). But as cloud capabilities expanded, MMS 2025 sessions showed an increasing confidence in cloud-only management. Features like Windows Autopilot (for zero-touch provisioning), Enterprise Patch Management via Intune (including the new Azure Update Management for servers), and Application deployment through Microsoft Store for Business successors were highlighted as robust enough to handle even large enterprises. Modern management now encompasses the full device lifecycle: from procurement (with hardware vendors supporting Autopilot out-of-box enrollment) to retirement (with remote wipe and deprovisioning). Senior engineers are sharing updated “best practices” to ensure cloud-managed endpoints remain compliant: e.g. making sure Conditional Access is tuned to not interfere with Autopilot enrollment (a lesson learned when too-strict MFA policies caused enrollment failures in some cases), and using compliance policies in Intune as a first-class control mechanism rather than traditional domain join GPOs.

Security in modern management is also a major focus. With devices off the corporate network, Zero Trust principles guide management: every device is treated as potentially external and must continually prove compliance. MMS 2025 reinforced using Entra ID (Azure AD) as the identity backbone for device trust, combined with Intune compliance checks and Microsoft Defender threat signals to govern access. The integration between Intune and security tools is tighter than ever – for example, Microsoft’s Endpoint Data Loss Prevention (DLP) can now receive signals from Intune about device risk posture, and Conditional Access can block access if a device falls out of compliance or if threat intelligence flags it. Attendees were urged to “regain control of your digital identity” (as one identity-focused session put it) by cleaning up old accounts and enforcing strong authentication, since identity is the new perimeter. In summary, modern management in 2025 means cloud-first, identity-driven management with continuous monitoring – a dynamic environment that demands automation and intelligence to manage effectively. That leads us to the next big update: the rise of AI and Copilots in the IT admin toolbox.

AI in the IT Toolbox: Copilots, Assistants, and Automation

Perhaps the most exciting development showcased at MMS 2025 was the integration of AI “Copilots” into IT operations. Microsoft and the community are bringing generative AI assistance to everything from endpoint management to security response. For those unfamiliar, Copilot is Microsoft’s branding for AI assistants powered by large language models (like GPT-4) that can understand natural language and generate responses or actions. At the conference, experts demonstrated how Copilots can help analyze data, automate tasks, and even write code or policy on behalf of IT pros. This is not just hype – concrete examples were in public preview.

Copilot in Intune: One of the headline announcements was that Microsoft Intune is getting its own Copilot integration. In preview, Intune’s portal now features a Copilot that can answer questions and perform actions related to device management. For instance, an admin can ask in natural language about the state of devices or policies. Currently, Copilot-assisted device queries let you investigate a single device in real-time or even query your entire fleet for a condition (e.g. “show all devices with BitLocker off”). The Copilot will translate that into the necessary filter or KQL (Kusto Query Language) behind the scenes and return results. Similarly, there’s Copilot assistance for device troubleshooting, which can identify issues on a device, compare its configuration with healthy peers, and explain error codes or remediation steps. This is a game-changer for helpdesk scenarios – instead of manually combing through event logs or Intune error messages, an engineer can simply ask the Copilot “why isn’t this device compliant?” and get an informed analysis.

The Intune Copilot doesn’t stop at queries. It’s also integrated with advanced management features. One preview feature shows Copilot working with Endpoint Privilege Management (EPM) – analyzing which applications are requesting elevated privileges and highlighting potential risks automatically. Another assists with policy configuration: the Copilot can summarize an Intune policy and assess its impact, or guide an admin on how to configure a certain setting. This kind of functionality is targeted at speeding up complex tasks – for example, summarizing a 10-page configuration profile into a few bullet points, or explaining what effect a certain combination of Conditional Access rules might have. Notably, Intune’s Copilot features a chat interface built into the portal, which retains context of previous questions and is optimized for administrative scenarios (each prompt starts fresh to avoid confusion, but the tool can gather context like the device or policy you’re viewing). Early feedback from the private preview (launched April 2025) influenced this design to ensure it “meets admins where they are” and simplifies complex tasks rather than adding clutter.

Windows Copilot and “Copilot+ PCs”: Another major topic was Windows Copilot and the new class of hardware labeled Copilot+ PCs. Windows Copilot (the AI assistant built into Windows 11) was introduced to end users in late 2023, but in 2024–2025 it has evolved, especially when paired with specialized hardware. Copilot+ PCs refer to Windows 11 machines that include a dedicated Neural Processing Unit (NPU), a chip designed to accelerate AI and machine learning tasks locally. These devices are Microsoft’s answer to AI capability on the edge: they promise faster local AI processing, better power efficiency for AI workloads, and new OS features collectively called “Windows AI Experiences.” Importantly for IT, a Copilot+ PC isn’t just about the chat bot in the taskbar – it unlocks a suite of AI-powered functionalities baked into Windows 11 when such hardware is present.

Some of the new Windows Experiences enabled by NPUs were highlighted at MMS. For example:

Recall – an AI-driven search that lets users find and “jump back into” anything they’ve seen on their PC before (documents, emails, web pages) using natural language. It’s like a personalized, local search engine for your work that understands queries like “find the PDF I was reading last week about zero trust.”
ClickToDo – an assistive feature that can identify text or UI elements on your screen and help you take action. For instance, you could press a shortcut and simply say “schedule a meeting next Friday” while an email is open – the AI will identify relevant text (like a date or topic in the email) and initiate the calendar appointment.
Cocreator – an image generation tool built into apps like Microsoft Paint, where the user can describe an image and have the AI create it. This showcases local generative AI for creativity without needing cloud services.
Generative Fill and Restyle in Photos – akin to Photoshop’s AI features, Windows can let you select part of an image and have the AI fill in the background or “reimagine” a photo in a different style.
Natural Language Windows Search – instead of typing exact filenames or keywords, users can describe what they’re looking for in plain English and Windows Search (enhanced by AI) will understand and find it. For example, “Show me the Excel file about Q3 budgets I edited in early July.”

These capabilities may sound very user-centric, but they have enterprise IT implications. First, such features require new hardware – NPUs with >40 TOPS (trillions of operations/sec) of AI performance – meaning hardware refresh plans should consider which laptops/desktops will support the productivity features of the future. Second, anything that indexes user data (like Recall indexing what a user has seen) raises compliance and privacy questions. The reassuring news from MMS 2025 is that Microsoft has built in administrative controls for all these AI features. In fact, on Copilot+ PCs most of these features are off by default until enabled by an administrator, and IT can manage them via policy. Windows 11 now includes a Policy CSP called “WindowsAI” that allows fine-grained control – e.g. an admin can allow or prohibit features like Recall, ClickToDo, Cocreator, generative image fill, etc., and even completely disable Windows Copilot on corporate devices. This means organizations can choose which AI-powered tools to empower their users with, and which to restrict for compliance. The presence of a “TurnOffWindowsCopilot” policy switch is evidence that Microsoft anticipates some enterprises will want the Copilot chat interface gone entirely, at least until it matures or can be governed. There was also discussion of data privacy – by keeping features on-device (especially with NPUs), data can be processed locally without streaming to the cloud, aligning with strict privacy requirements. This addresses one of the early concerns: will Copilot send our confidential data to Azure/OpenAI? – for these Windows Experiences, much of the processing can be done locally, and if cloud AI is used, Microsoft has pledged it won’t use customer data to train public models.

From a senior IT perspective, a key takeaway is “you can manage it.” We don’t have to fear these new AI capabilities spilling into our environment unmanaged. The conference showcased Intune’s ability to toggle these features via configuration profiles. For example, if Recall is deemed too risky (since it stores a history of user content on the device), IT can simply not enable it – or set policies to control how much data Recall keeps and for how long. Conversely, if a team sees value in, say, AI-powered image generation for marketing users but not for developers, those settings can be targeted to specific groups. The enterprise management story for Copilot+ PCs is still evolving (some of these CSPs and Intune settings were “still in development” as of MMS 2025), but the groundwork is clearly there to adopt AI in a controlled, compliant manner.

What about the practical benefits? Beyond the glitz of AI features, MMS speakers urged technical decision-makers to consider why and where to leverage these tools. One session cut through the hype by asking: “What does it do for me as an admin? Does it solve my problems or just add noise?”. The consensus was that AI co-pilots can significantly augment IT work if used wisely. Imagine slashing helpdesk ticket triage time with local AI analyzing issues, or spotting log anomalies across thousands of devices before they become major incidents. These aren’t pipe dreams – with proper data, an AI can correlate events or error patterns much faster than a human. Microsoft has even demonstrated prototypes where an Intune Copilot might proactively alert on “devices with abnormal login failure rates” or flag a policy that is causing devices to crash. The speed and scale of AI are its strengths: routine tasks like generating a PowerShell script to, say, remove bloatware can be done by a Copilot in seconds, and complex pattern recognition (like finding which PCs might be impacted by a faulty update) can happen in the background continuously. AI can also bring natural language convenience to IT. Rather than remembering a slew of admin portals and PowerShell cmdlets, an IT specialist can ask, “Give me a report of all Windows 11 laptops that haven’t installed the latest security patch” and the Copilot (with proper permissions) can fetch or generate that report. This allows IT staff to focus more on decision-making and less on plumbing.

Security Copilot: AI as a Security Analyst

No discussion of AI in IT would be complete without mentioning Security Copilot – Microsoft’s AI assistant for cybersecurity. While not the primary focus of this blog, it’s worth noting how security operations align with modern management. Security Copilot, introduced in 2024, is built on GPT-4 and infused with security-specific knowledge. MMS 2025 sessions demonstrated how it can analyze incidents, summarize threats, and even provide step-by-step guidance during an incident response. What’s intriguing for endpoint management teams is the concept of Security Copilot “AI Agents.” These are like mini AI workflows that can be triggered by specific events or tasks. For example, beyond the generic Q&A chatbot, Microsoft is developing specialized agents – one out-of-the-box agent helps triage user-reported phishing emails automatically. In the future, you could see agents for Intune deployment health checks or inventory risk analysis, as security and management data converge. Imagine an “Intune Deployment Health” AI agent that continuously watches your device enrollments, configurations, and updates rollout – it could alert you if a new policy you set is causing failures, or if a subset of devices has significantly lower compliance rates than others, and even suggest what the root cause might be. Similarly, an “Inventory Risk Analyzer” could look at your device inventory against known vulnerabilities (e.g. OS versions, missing patches, insecure configurations) and highlight the most at-risk endpoints proactively. This level of intelligent monitoring could revolutionize how we approach preventive maintenance and security in endpoint management. Rather than waiting for an audit or a pen-test report, the AI would constantly audit configurations and highlight weaknesses.

The integration of Intune with security tools was another point reinforced at the summit. Compliance data from Intune can feed into Microsoft Defender and Sentinel. For example, a Defender XDR (extended detection & response) system can query Intune for device risk level and take action like isolating a device if it’s non-compliant and showing signs of compromise. With AI in the mix, these decisions could be automated in more nuanced ways (“only isolate if non-compliance is related to missing critical updates and we see active exploitation of that vulnerability”). Security Copilot could serve as the brain coordinating these signals – truly embodying Microsoft’s vision of holistic, AI-driven security across Intune, Entra ID, Defender, and more. For IT managers, this means that silos between endpoint management and security operations are blurring: modern management solutions are part of the security fabric, and AI is amplifying that connection. It underscores the need to have cross-discipline knowledge – endpoint admins should understand security fundamentals, and security analysts should be aware of Intune configurations – because the AI will be looking at all of it in tandem.

Preparing Your IT Team for the AI-Driven Future

The advancements shared at MMS 2025 make one thing clear: the future of IT management is AI-augmented and cloud-powered. For senior IT engineers and technical managers, the challenge now is guiding your teams through this transition. Here are a few concrete steps and considerations to keep in mind:

Stay Informed and Hands-On: The pace of innovation is high. Encourage your team to follow update channels and preview programs for tools like Intune and Security Copilot. The features in public preview (like Intune’s Copilot and Windows AI features) will likely become general availability features within the next year. Allocating time for engineers to test these in a lab or pilot environment is critical. MMS 2025 itself, with its level 300–500 deep dives, showed that those who experiment early can provide valuable feedback and shape the tools. For example, early testers of Intune’s Copilot in private preview influenced its design towards more “open prompting” after finding the initial scoped Q&A too limiting. Being part of that conversation ensures your organization’s needs are considered.
Develop New Skills in AI and Data Analysis: Managing AI-powered tools requires understanding how they work and what their limitations are. Train your staff not just on how to click the Copilot button, but also on the basics of AI model behavior. There will be times when the Copilot is confidently wrong or when a prompt needs refining. Administrators should learn prompt engineering basics – e.g. how to phrase a question to get the best report – and how to verify AI outputs. Additionally, skills in KQL for Intune’s advanced queries or familiarity with Azure OpenAI Service (if building custom internal copilots) could be very valuable. These might sound more like data science than IT, but the roles are converging. As one speaker put it bluntly: “AI is not just another tool – it’s fundamentally changing how IT is done”. Embracing that mindset is key.
Update Policies and Governance: With great power comes great responsibility. Review your IT policies to account for AI-driven features. This includes drafting guidelines for acceptable use of generative AI at work (e.g. should employees use Windows Copilot to compose emails? Are there data sensitivity concerns?), and updating security/privacy policies to cover new data that might be stored (e.g. Copilot “Recall” snapshots on devices). Leverage the management controls available: for instance, if your company is not ready for a certain feature, use Intune’s configuration profiles to disable it at the outset. On the flip side, if you see a feature that could boost productivity safely, advocate for its enablement and provide training on it. A modern IT leader should create a balance where innovation isn’t stifled by fear, but guardrails are firmly in place. Microsoft’s Responsible AI principles (fairness, privacy, security, etc.) are built into their Copilot products – make sure your implementation of these tools adheres to the same ethos.
Plan Hardware Refresh Strategically: Given the emergence of NPU-equipped PCs, your hardware purchasing strategy may need revision. When budgeting for new laptops/desktops, consider opting for models that meet the Copilot+ PC specs (e.g. >=16 GB RAM, >=256 GB SSD, and an NPU with 40+ TOPS performance). Even if you don’t intend to enable all AI features immediately, these specs also generally correlate with longer device lifespans and better performance, which is a win for users and IT. Make an inventory of your current endpoints to identify how many are “AI-capable.” MMS experts suggested doing this inventory now: “You can’t manage what you can’t see – know which devices have NPUs for targeting policies, planning upgrades, and troubleshooting”. Intune’s reporting and hardware inventory can be extended (via scripts or custom attributes) to mark NPU presence, and there are quick local checks (Task Manager now shows an NPU graph if one is present) that teams can use for spot checks. Armed with this information, you can prioritize early adopters for AI features or conversely ensure non-capable devices get an alternate configuration.
Foster a Culture of Automation and Innovation: Finally, encourage your IT staff to offload repetitive tasks to AI and automation. Free up humans for higher-level planning and creative problem solving. If an engineer spends hours every month writing similar PowerShell scripts or combing through device logs, challenge them to see if an AI tool can handle the first draft. Similarly, integrate automation pipelines (DevOps workflows, Infrastructure as Code) with these AI capabilities. For example, a script generated by Copilot might automate a task, but you’ll want it checked into source control and reviewed. AI can jump-start work, but human oversight and continuous improvement turn those suggestions into robust solutions. Mentor your team to use AI as a partner, not a crutch – it can supercharge productivity, but it’s most effective in the hands of knowledgeable professionals who guide it.

Conclusion: Leading the Way Forward

The IT landscape in 2025 is both exciting and challenging. Cloud-based modern management has delivered flexibility and scale, but also introduced continuous change and complexity. Now, AI and Copilot technologies are rising to tame that complexity, offering a new layer of intelligence that can operate at machine speed. As highlighted at MMS 2025, we stand at an inflection point where IT management is becoming less about manually wrangling systems and more about supervising intelligent platforms that do the heavy lifting. Senior IT leaders should seize this moment to push their organizations forward: adopt modern management fully, experiment with AI copilots in safe environments, and develop the skills and policies needed to harness these tools responsibly.

In doing so, you’ll transform your IT department from a reactive service desk into a proactive innovation center. Routine issues get solved faster (or prevented entirely) with AI-assisted insights, users get more out of their technology (with features like Windows Copilot empowering them in their flow of work), and IT professionals can focus on strategic initiatives rather than firefighting. The core mission of IT remains the same – to deliver a secure, efficient, and empowering tech environment for the organization – but the means to that end are evolving. Embracing cloud and AI is now key to that mission.

To conclude, take inspiration from the experts and peers at events like MMS: share knowledge, stay curious, and don’t be afraid to pilot new approaches. The next few years will undoubtedly bring even more advanced capabilities (today’s “preview” is tomorrow’s standard). By laying a strong foundation now in modern management practices and cultivating AI fluency in your team, you’ll ensure your organization is not just adapting to the future of IT, but actively shaping it. The age of AI-enhanced, modern management is here – and it’s an exciting time to be an IT leader on the frontlines of this transformation.

Sources: The insights and examples above are drawn from MMS 2025 sessions and materials, including “Troubleshooting Tips for the Intune Modern Managed Client”, “Copilot in Intune: Navigating the Future of Secure Endpoint Management”, “Cutting Through the Hype: Copilot+ PCs’ Impact on Sysadmin Work”, and “Securing Tomorrow: Unveiling the Power of Microsoft Security”, among others. These illustrate the current state of enterprise IT and Microsoft’s roadmap, as discussed by community experts and Microsoft product engineers at the summit. All signs indicate that AI and cloud-driven management will define the next chapter of IT – so let’s make sure we’re ready for it.