You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: learn-pr/device-partner-university/how-windows-enables-ai-agents/includes/1-introduction.md
+4-4Lines changed: 4 additions & 4 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -1,12 +1,12 @@
1
-
AI is no longer confined to individual apps — it’s becoming part of the operating system itself. Windows is evolving from a traditional OS into a foundation for intelligent assistants, agents, and AI-powered workflows that help people work more efficiently.
1
+
AI is no longer confined to individual apps—it’s becoming part of the operating system itself. Windows is evolving from a traditional OS into a foundation for intelligent assistants, agents, and AI-powered workflows that help people work more efficiently.
2
2
3
-
AI has traditionally shown up as something users deliberately launched—through a chatbot, a webpage, or a standalone application. Now that’s changing. Intelligence is being built directly into Windows — into its shell, into system experiences, and into everyday workflows. Instead of being “one more app to open,” AI becomes part of how people use their PCs.
3
+
AI has traditionally shown up as something users deliberately launched—through a chatbot, a webpage, or a standalone application. Now that’s changing. Intelligence is being built directly into Windows—into its shell, into system experiences, and into everyday workflows. Instead of being "one more app to open," AI becomes part of how people use their PCs.
4
4
5
-
In this module, you’ll explore how Windows is becoming the canvas for AI by integrating intelligence across the operating system, silicon, and hardware. The goal is to help organizations move from experimentation to execution.
5
+
In this module, you explore how Windows is becoming the canvas for AI by integrating intelligence across the operating system, silicon, and hardware. The goal is to help organizations move from experimentation to execution.
6
6
7
7
## Learning objectives
8
8
9
-
By the end of this module, you’ll be able to:
9
+
By the end of this module, you're able to:
10
10
11
11
- Explain how AI, agents, and Microsoft 365 Copilot are integrated into familiar Windows experiences
12
12
- Describe how Copilot+ PCs support advanced on-device AI workloads
Copy file name to clipboardExpand all lines: learn-pr/device-partner-university/how-windows-enables-ai-agents/includes/2-why-windows-is-evolving-as-platform-ai-and-agents.md
+9-9Lines changed: 9 additions & 9 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -35,13 +35,13 @@ Instead of step-by-step instructions from IT, policies and apps are automaticall
35
35
36
36
Rather than guessing or searching online, an agent can gather logs, summarize findings, suggest actions, and when permitted, carry out remediation. That means fewer remote sessions and less time in Event Viewer.
37
37
38
-
The value is not “AI for AI’s sake.” The value is **fewer repetitive tickets, faster remediation, and more consistent user experiences**—while IT maintains control through policy and governance.
38
+
The value isn't “AI for AI’s sake.” The value is **fewer repetitive tickets, faster remediation, and more consistent user experiences**—while IT maintains control through policy and governance.
39
39
40
40
## What AI agents are
41
41
42
42
Before going deeper into how agents show up in Windows, it helps to define what we mean by an **AI agent**.
43
43
44
-
An AI agent is an AI-powered tool that can **do work on your behalf**, not just answer questions. While [Microsoft 365 Copilot](/copilot/microsoft-365/microsoft-365-copilot-overview) focuses on conversation, guidance, and suggestions, **agents are designed to take action** — always within the permissions and policies your organization sets.
44
+
An AI agent is an AI-powered tool that can **do work on your behalf**, not just answer questions. While [Microsoft 365 Copilot](/copilot/microsoft-365/microsoft-365-copilot-overview) focuses on conversation, guidance, and suggestions, **agents are designed to take action**—always within the permissions and policies your organization sets.
45
45
46
46
Put simply:
47
47
@@ -50,7 +50,7 @@ Put simply:
50
50
51
51
## How to think about agents
52
52
53
-
You can think of an AI agent as a **digital teammate**. It can understand intent, reason through multi-step tasks, and take action across apps and services — all while respecting identity, permissions, and organizational policy. Instead of completing every individual step yourself, you describe the outcome you want, and the agent helps carry it out.
53
+
You can think of an AI agent as a **digital teammate**. It can understand intent, reason through multi-step tasks, and take action across apps and services—all while respecting identity, permissions, and organizational policy. Instead of completing every individual step yourself, you describe the outcome you want, and the agent helps carry it out.
54
54
55
55
## What agents can assist with
56
56
@@ -62,7 +62,7 @@ For example, an AI agent could:
62
62
-**Summarize files or folders** to help prepare a brief or report
63
63
-**Guide users through setup or troubleshooting** without a ticket right away
64
64
65
-
In other words, agents don’t just provide answers — they take action under organizational policy, identity, and consent.
65
+
In other words, agents don’t just provide answers—they take action under organizational policy, identity, and consent.
66
66
67
67
## Evolving the Windows platform
68
68
@@ -78,7 +78,7 @@ This shift means agents aren’t running “off to the side.” They become firs
78
78
79
79
## Phases of AI evolution in Windows
80
80
81
-
:::image type="content" source="../media/evolution-of-ai.svg" alt-text="Three phases of AI evolution showing a human with an assistant, human-led agents, and human-led agent-operated workflows.":::
81
+
:::image type="content" source="../media/evolution-of-ai.svg" alt-text="Infographic showing three phases of AI evolution showing a human with an assistant, human-led agents, and human-led agent-operated workflows.":::
82
82
83
83
| Phase | Description | What it looks like in practice |
- Move on to other tasks while your agents work for you
98
98
99
99
> [!IMPORTANT]
100
-
> Features and capabilities in this section **labeled as preview**are not final and may change before general availability. Features, timelines, and experiences are subject to change.
100
+
> Features and capabilities in this section **labeled as preview**aren't final and may change before general availability. Features, timelines, and experiences are subject to change.
101
101
102
102
## A more secure and resilient foundation
103
103
@@ -118,7 +118,7 @@ Most IT teams recognize the patterns immediately. A user resets their VPN creden
118
118
119
119
Traditionally, these situations turn into tickets, remote sessions, and manual investigation. Users search online, wait in queues, and IT walks through the same troubleshooting steps over and over again.
120
120
121
-
With AI and agent capabilities built directly into Windows, these aren’t simply “support incidents” anymore — they become guided workflows. The device itself can help users diagnose issues, gather relevant context for IT, or even resolve common problems automatically before they hit the help desk.
121
+
With AI and agent capabilities built directly into Windows, these aren’t simply “support incidents” anymore—they become guided workflows. The device itself can help users diagnose issues, gather relevant context for IT, or even resolve common problems automatically before they hit the help desk.
122
122
123
123
:::image type="content" source="../media/ai-decision-flow.svg" alt-text="Infographic showing traditional AI workflow versus integrated AI in Windows.":::
124
124
@@ -130,7 +130,7 @@ With AI and agent capabilities built directly into Windows, these aren’t simpl
130
130
| Failed update | Downtime → troubleshooting → sometimes reimage | Recovery guidance with automated rollback options |
131
131
| “What changed recently?” | IT manually checks policy, app installs, and updates | System surfaces recent changes automatically as context |
132
132
133
-
For example, a salesperson is about to join a customer call and their VPN connection fails. Instead of submitting a ticket and waiting, Windows detects repeated connection errors, launches guided troubleshooting, repairs the profile, and verifies connectivity. The user joins the meeting on time — IT sees the summary afterward instead of an urgent escalation.
133
+
For example, a salesperson is about to join a customer call and their VPN connection fails. Instead of submitting a ticket and waiting, Windows detects repeated connection errors, launches guided troubleshooting, repairs the profile, and verifies connectivity. The user joins the meeting on time—IT sees the summary afterward instead of an urgent escalation.
134
134
135
135
That’s the difference: not replacing IT but moving routine fixes closer to the user while preserving auditability and control.
136
136
@@ -166,4 +166,4 @@ List three ways AI agents integrated into Windows could reduce repetitive suppor
166
166
167
167
Windows is evolving to embed AI into the operating system, reducing friction and enabling smarter workflows. This shift sets the stage for bringing AI and agents into familiar Windows experiences.
168
168
169
-
Next, we’ll look at **how AI and agents show up in everyday Windows experiences** and what it means to bring assistance directly into familiar surfaces such as the taskbar, File Explorer, and system settings.
169
+
Next, we look at **how AI and agents show up in everyday Windows experiences** and what it means to bring assistance directly into familiar surfaces such as the taskbar, File Explorer, and system settings.
Copy file name to clipboardExpand all lines: learn-pr/device-partner-university/how-windows-enables-ai-agents/includes/3-ai-and-agent-integration-across-windows-experiences.md
+4-4Lines changed: 4 additions & 4 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -44,7 +44,7 @@ These experiences are designed to feel incremental, not disruptive.
44
44
-**Voice interaction with Microsoft 365 Copilot**
45
45
Users can brainstorm ideas, draft responses, or prepare meeting content using voice-reducing app switching and manual typing.
46
46
47
-
Together, these capabilities make AI feel like a natural extension of Windows instead of “one more app to launch.”
47
+
Together, these capabilities make AI feel like a natural extension of Windows instead of "one more app to launch."
48
48
49
49
> [!IMPORTANT]
50
50
> These AI experiences within Windows—such as Microsoft 365 Copilot File Explorer integration, and voice interactions—are **in preview** and may change.
@@ -64,7 +64,7 @@ For example:
64
64
-**Research and troubleshooting in Edge for Business**
65
65
AI capabilities in [Microsoft Edge for Business](https://learn.microsoft.com/en-us/deployedge/microsoft-edge-for-business) help synthesize information across multiple tabs while maintaining enterprise security, compliance, and control.
66
66
67
-
For end users, these same experiences reduce common support requests—such as “where do I find this?” or “how do I get started?”—before a ticket is ever created.
67
+
For end users, these same experiences reduce common support requests—such as "where do I find this?" or "how do I get started?"—before a ticket is ever created.
68
68
69
69
## Evolving the Windows platform for agents
70
70
@@ -75,7 +75,7 @@ Windows includes **native agent infrastructure** designed to help agents operate
75
75
By embedding agent capabilities into core Windows services:
76
76
77
77
- Copilot and other agents can coordinate tasks across apps, files, and system functions
78
-
- Users receive context-aware assistance based on what they are doing
78
+
- Users receive context-aware assistance based on what they're doing
79
79
- IT teams retain control over what agents are allowed to do and when
80
80
81
81
Agents become first-class participants in the Windows experience, operating under the same governance and security expectations as other system components.
@@ -88,7 +88,7 @@ All of the AI and agent experiences discussed in this module run on a native AI
88
88
89
89
Foundry supports the full AI development lifecycle—from model selection to deployment—and enables workloads to run across **CPUs, GPUs, and NPUs**. This flexibility allows organizations to choose the appropriate execution path based on performance, privacy, and operational requirements.
90
90
91
-
For IT teams, this means AI capabilities can be introduced into **line-of-business applications** using a standardized platform that aligns with Windows security, identity, and management controls. Rather than treating AI as a separate environment, Foundry integrates AI directly into Windows-based workflows in a way that is predictable, supportable, and scalable.
91
+
For IT teams, this means AI capabilities can be introduced into **line-of-business applications** using a standardized platform that aligns with Windows security, identity, and management controls. Rather than treating AI as a separate environment, Foundry integrates AI directly into Windows-based workflows in a way that's predictable, supportable, and scalable.
92
92
93
93
## How Microsoft Foundry on Windows supports IT and developers
Copy file name to clipboardExpand all lines: learn-pr/device-partner-university/how-windows-enables-ai-agents/includes/4-device-based-ai-capabilities-copilot-plus-pcs.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -12,7 +12,7 @@ Together, these factors help AI move from something that feels experimental to s
12
12
13
13
## Understanding the hardware difference
14
14
15
-
Copilot+ PCs are not simply faster versions of existing devices. They use a different balance of compute resources, with each processor type handling a specific kind of workload.
15
+
Copilot+ PCs aren't faster versions of existing devices. They use a different balance of computer resources, with each processor type handling a specific kind of workload.
16
16
17
17
:::image type="content" source="../media/understanding-processors-for-ai.svg" alt-text="Infographic titled 'Understanding Processors for AI' comparing CPU, GPU, and NPU roles. The CPU handles general-purpose computing but isn’t optimized for sustained AI workloads. The GPU accelerates graphics and some AI tasks through parallel processing but can be power‑intensive. The NPU provides efficient, low‑power on‑device AI inference using specialized hardware.":::
0 commit comments