Scaling AI and automation in your organization must take security into account. As more businesses take advantage of advanced technology, like AI agents, gaps in security or data can create vulnerabilities and inefficiencies. 

Microsoft Ignite 2025 focused on four major themes: AI, security, productivity and governance. These updates provide more advanced digital transformation options for businesses.  

As your organization develops its technology roadmap for 2026 and beyond, consider how the following services could impact your business. 

Microsoft Defender and Purview  

Microsoft is improving security across its products but is especially focused on AI.  

If hackers gain access to a Copilot license, they can build an AI agent. The agent will act on the hacker’s behalf, creating a “shadow AI” scenario.  

Microsoft Defender and Purview will now secure AI agents as first-class assets. This helps Microsoft build a security framework around these agents. 

Microsoft is also adding Predictive Shielding and proactive threat disruption in Defender. AI-based features bring proactive enhancements to Defender, including: 

• Broadened security footprint 

• Improved endpoint protections across all products 

• Dynamic restricted access to sensitive data 

• Faster response time for security teams  

• Minimized operational overhead 

Why it Matters: 

• Small IT teams can benefit from simplified endpoint security and compliance 

• Securing AI agents as a first-class asset helps reduce the risk of shadow AI while also strengthening Zero Trust posture.  

Your Next Steps: 

• Enable Intune Suite features for unified endpoint management (UEM).  

• Apply Purview policies for agent governance and compliance.  

Power Platform and Copilot Studio 

Several improvements are in store for Power Platform and Copilot Studio. Enhanced reporting for apps, workflows and agents brings more insight into how users interact and use AI agents. Other updates help make app-building more accessible to teams of all experience levels:  

Vibe Power Apps: A conversational app-builder. The no-code/low-code program allows users to describe their ideal app to the AI, using natural language, and it generates code for the app automatically.  

Model Context Protocol (MCP): A protocol that standardizes the way AI integrates and shares data with external systems. This protocol enables secure two-way connections without the need to manually connect to APIs for complicated connections.  

Why it Matters: 

• Non-technical staff are empowered to quickly build apps themselves. 

• IT teams benefit from improved governance and visibility into user behavior.  

Your Next Steps: 

• Try Copilot Studio for internal automation. 

• Use MCP to streamline secure workflows. 

Azure Fabric IQ and Foundry IQ 

AI runs on data, but complex or unorganized data can make AI less effective. Microsoft made updates that unify data layers while enabling AI agents to act on governed data for insights and automation.  

Azure Fabric IQ and Foundry IQ: These programs unify data across different platforms using a shared intelligence layer. This produces a more cohesive foundation for enterprise AI agents by making data more easily accessible.  

OneLake: OneLake breaks down data silos, making complex data accessible for Power BI and Copilot. It creates a single, unified data lake for your entire organization. Described as the OneDrive for data, this data storage app is automatically available with a Fabric tenant. 

Why it Matters: 

• Consistent, secure data helps AI improve its decision-making  

• AI architecture is future-proofed for scaling automation 

Your Next Steps: 

• Consolidate your organization’s data with OneLake for AI readiness and analytics. 

• Explore Foundry IQ to experience agent-driven workflows.  

Copilot  

Copilot will be updated to be more proactive and improve its creative outputs. Currently, Copilot is reactive, only assisting after being prompted. Work IQ changes how end users interact with Copilot across applications.  

With Work IQ, Copilot learns about the user through chats, emails, documents and more. Using this information, Copilot builds a “working memory” of the user. It may make suggestions or add additional information before being prompted.  

For example: a user has a meeting on their calendar for later in the day. Copilot uses chat logs and emails to learn what the meeting is about. Before the meeting, Copilot autonomously prepares materials and alerts the user.  

Dedicated agents will also be available in Word, Excel and PowerPoint to enable iterative content creation and chat-based management.  

Why it Matters: 

• Upgrades to Copilot accelerate document and data workflows 

• More proactive AI reduces the need for manual effort and improves the quality of its outputs via iterative automation 

Your Next Steps: 

• Train teams on the capabilities of the new Copilot agents. 

• Use iterative workflows that enable faster content development. 

Agent 365  

Agent 365 is a centralized management and governance platform for AI agents. IT staff can use it to develop and deploy AI agents across an enterprise. It also provides much-needed transparency into users’ interactions with the agents they created through Copilot.  

Identity and access controls are also available via Entra ID, enabling telemetry and compliance monitoring.  

Why it Matters: 

• The added visibility helps prevent shadow AI risks and ensures secure adoption by users 

• IT teams benefit from simplified governance  

Your Next Steps: 

• Implement Agent 365 early to manage the complete AI lifecycle. 

• Before scaling agent deployments, define your governance policies.  

Anders Technology is a Microsoft Solutions Partner with Advanced Specialization in Azure Virtual Desktops. Our experience helps clients optimize their Microsoft suite of products. Learn more about our services, and the associated cost, by requesting a meeting below.