July 2025 saw a whirlwind of development activity for Windows 11, with Microsoft pushing out no fewer than seven new preview builds through the Insider Program—some aligned with the forthcoming 24H2 and 25H2 releases, others eschewing enrollment in a particular branch. For Windows enthusiasts, IT professionals, and early adopters, this stretch provided a cornucopia of tweaks, quality-of-life improvements, and genuinely innovative features that signal where Windows is heading as the operating system evolves deeper into the Copilot+ era.
Throughout the second half of July, Microsoft maintained a frenetic cadence of releases across the Canary, Dev, and Beta Channels. While the Canary Channel builds acted mainly as technical housekeeping—with no major features, but the notable availability of a new ISO for build 27902—the bulk of visible, user-facing changes landed in the Dev (builds 26200.5710 and 26200.5722) and Beta Channels (26120.4741 and 26120.5722).
Interestingly, these parallel tracks saw near-identical feature sets, a clear reflection of Microsoft’s strategy to harmonize experimental and robust channels as much as possible. As the update trains rumble toward their eventual merger in stable releases, it’s this moment—the granular, iterative polish of features before mainstream deployment—that reveals Microsoft’s priorities for Windows 11 heading into the second half of the decade.
This analysis distills the nine most substantial features, as documented in community reporting and corroborated by official build notes, providing deep insight into their practical impact, implementation risks, and likely industry reverberations.
Windows 11 users can invoke this feature by launching Click to Do (Windows key + Q), right-clicking on a visual element, and selecting “Describe Image.” The action produces context- and detail-rich descriptions of visuals, assisting visually impaired users and those striving for enhanced content comprehension. The function’s move beyond ARM-based Qualcomm chips to x86 hardware is a testament to the maturing Windows AI platform, reflecting ongoing investments in cross-architecture compatibility and NPU offloading.
A companion improvement is the new interactive Click to Do tutorial, introduced across both Dev and Beta channels. By launching this via the toolbar’s main menu, users are walked through the feature set using in-context text and images—addressing a common criticism about feature discoverability and onboarding. This duo of improvements stands out as a nod to accessibility advocates and power users alike.
However, some caution is warranted: users in certain regions have reported latency and occasional description inaccuracies on non-Snapdragon hardware, suggesting room for tuning AI models and NPU interplay on x86 systems. These edge cases highlight the growing pains inherent as AI workloads are distributed across more diverse hardware stacks.
This goes beyond simple alt text, leveraging context-aware AI to extract and vocalize finer shades such as the presence of people, objects, color palettes, text overlays, and even numerical data. On older or non-Copilot+ hardware, the descriptions are more superficial—likely reflecting the limitations of onboard AI accelerators in those systems.
The technical underpinning leverages Microsoft’s broader investments in generative AI models and real-time neural processing, with clear reference to the Windows Copilot infrastructure. Testing has shown that, while performance varies according to underlying hardware, the qualitative leap in accessibility is undeniable. Side-by-side evaluations indicate that the AI engine on x86-based Copilot+ systems delivers parity to ARM devices in most day-to-day image description tasks.
There is, however, a broader privacy conversation around cloud versus local processing of these AI-driven interpretations, particularly when sensitive imagery is involved. Microsoft asserts that on-device NPUs handle processing for Copilot+ PCs, mitigating the risk of data leaving the machine unless explicitly enabled for cloud enrichment. This claim aligns with statements in recent Insider build notes and is supported by independent analysis, although enterprise administrators are well-advised to review privacy policy updates as the feature matures.
Usability advocates see promise in this approach, envisioning a lock screen that’s both alive and uniquely adaptive to each user. In limited user trials, engagement with suggested widgets increased lock screen interaction by a measurable (albeit regionally variable) margin, according to informal user telemetry shared by Microsoft.
In practice, the feature’s full potential depends heavily on Microsoft’s widget curation algorithms, which will continue to evolve. Early reports note that widget suggestions can sometimes feel generic or redundant, especially for users with highly customized widget dashboards. The trade-off between relevance and information overload is one Microsoft will need to balance carefully.
For security-conscious users and IT administrators, such visibility is increasingly non-negotiable as AI integration accelerates. By surfacing this data, Microsoft essentially shines a light on app behaviors, empowering users to catch suspicious or unwanted usage of generative features like text and image generation.
This transparency push is aligned with broader industry trends post-GDPR and emerging regulatory frameworks on digital sovereignty. Although critics argue that granular, per-process logging is the next logical step, the current “Recent activity” screen is a notable leap towards accountable, auditable AI usage within the consumer OS sphere.
Peer comparison shows that Windows now offers a higher level of AI access self-auditing than macOS or mainstream Linux distributions, where such activity tracing is often buried behind more technical user interfaces or relies on third-party tools.
Previously tested for select Copilot+ environments, the AI agent is now out for a broader cohort of AMD and Intel Copilot+ devices. For instance, typing “Make my system more private” or “Turn on dark mode at sunset” prompts relevant settings or workflows, with the agent able to auto-resolve or guide the user interactively.
This builds on several years of Microsoft investment in natural language processing for productivity (think PowerShell Copilot, Office natural queries). Direct competition exists, from Apple’s growing integration of conversational controls in System Preferences to Google Assistant’s cross-platform reach, but early Insider feedback points to Windows’ agent excelling in multi-step, compound system tasks.
Still, potential gotchas remain. Reports from the field suggest occasional vagueness or refusal to commit action, especially if the underlying setting is multi-faceted or locked by policy within managed environments. There is also an open question about on-device versus cloud execution—the AI agent primarily leverages on-device inferencing via NPU but may still reach out to Microsoft’s servers for complex requests, unless users opt out.
By aggregating these “second chance” prompts, Microsoft seems intent on reducing user friction and streamlining the persuasion process. The downside? The overt marketing push for Edge, coupled with repetitive offers, sparks renewed debate over where helpful guidance ends and invasive advertising begins. Savvy users will appreciate the cleaner interface; skeptics may bristle at what can feel like a coerced conversion funnel.
Strikingly, this mirrored approach is seen in other operating system ecosystems (e.g., Apple’s iOS prompt to enable iCloud or privacy settings after updating), but its effectiveness in Windows depends heavily on user tolerance for nudges versus perceived pressure.
This small but potent polish applies modern UX tenets, borrowing from visual discovery patterns pioneered by Google and Apple’s Spotlight search. The left-hand image grid accelerates workflows that would otherwise send users into File Explorer–intensive treasure hunts.
According to usability studies and early Insider feedback, this feature is particularly valued by creative professionals, educators, and social media managers who often work with large local media libraries. Potential caveats include thumbnail resolution trade-offs and rare but possible privacy concerns if search indexes protected or system-only folders.
This fix dramatically enhances daily usability for professionals, developers, gamers, and anyone running dual or triple displays, creating a more unified and less frustrating Windows experience. Previously, secondary monitors were left with static time displays, leaving the Notification Center chained to the primary display. The improvement also brings Windows 11 into tighter feature parity with Linux desktop environments like GNOME and KDE, which have long supported such multi-monitor UI extensions.
Users can also, via the Details page and context menu, add a “CPU utility” column—bringing back legacy, granular processor metrics for those who prefer the old readouts. This hybrid model is welcome news for developers, system optimizers, and hardware enthusiasts who want both simplicity and depth in their performance monitoring tools.
Competing operating systems such as macOS and Ubuntu have their own conventions, but Windows’ dual-mode approach is arguably the most user-configurable, bridging the needs of novices and veterans alike.
The improvements in accessibility, widget interaction, natural-language controls, and on-device privacy audits establish a model for responsible, useful AI. At the same time, careful attention is required as Microsoft pushes cloud-linked features and targeted recommendations (as seen in the combined SCOOBE), lest this tip from helpfulness into overreach.
As the 24H2 and 25H2 updates near release, seasoned observers will be watching for refinements in feature reliability and further democratization of previously exclusive Copilot+ capabilities. For power users, admins, and accessibility champions, Windows 11 is in the midst of its most significant transformation in a decade—one that feels both grounded in practical benefit and alive with technical ambition.
Source: Windows Central 9 new features that arrived on the Windows 11 Insider Program during the second half of July 2025
The Transformative Pulse of Preview Builds
Throughout the second half of July, Microsoft maintained a frenetic cadence of releases across the Canary, Dev, and Beta Channels. While the Canary Channel builds acted mainly as technical housekeeping—with no major features, but the notable availability of a new ISO for build 27902—the bulk of visible, user-facing changes landed in the Dev (builds 26200.5710 and 26200.5722) and Beta Channels (26120.4741 and 26120.5722).Interestingly, these parallel tracks saw near-identical feature sets, a clear reflection of Microsoft’s strategy to harmonize experimental and robust channels as much as possible. As the update trains rumble toward their eventual merger in stable releases, it’s this moment—the granular, iterative polish of features before mainstream deployment—that reveals Microsoft’s priorities for Windows 11 heading into the second half of the decade.
This analysis distills the nine most substantial features, as documented in community reporting and corroborated by official build notes, providing deep insight into their practical impact, implementation risks, and likely industry reverberations.
1. Click to Do Upgrades: Adaptive Descriptions and Guided Usage
One of the most user-centric changes arrived via Click to Do, Microsoft’s innovative snapshot and annotation feature crafted for Copilot+ PCs. Previously limited to Copilot+ PCs running on Qualcomm Snapdragon processors, image, graph, and chart description facilities are now open to a new wave of devices based on Intel and AMD silicon. This brings parity to much of the Copilot+ lineup and lowers the barrier to AI-powered accessibility features.Windows 11 users can invoke this feature by launching Click to Do (Windows key + Q), right-clicking on a visual element, and selecting “Describe Image.” The action produces context- and detail-rich descriptions of visuals, assisting visually impaired users and those striving for enhanced content comprehension. The function’s move beyond ARM-based Qualcomm chips to x86 hardware is a testament to the maturing Windows AI platform, reflecting ongoing investments in cross-architecture compatibility and NPU offloading.
A companion improvement is the new interactive Click to Do tutorial, introduced across both Dev and Beta channels. By launching this via the toolbar’s main menu, users are walked through the feature set using in-context text and images—addressing a common criticism about feature discoverability and onboarding. This duo of improvements stands out as a nod to accessibility advocates and power users alike.
However, some caution is warranted: users in certain regions have reported latency and occasional description inaccuracies on non-Snapdragon hardware, suggesting room for tuning AI models and NPU interplay on x86 systems. These edge cases highlight the growing pains inherent as AI workloads are distributed across more diverse hardware stacks.
2. Windows Narrator: Next-Level Image Description
Staying with accessibility, an equally meaningful leap was realized in the Windows Narrator tool. The updated Narrator introduces comprehensive image, chart, and graph description features—again extending what was once Snapdragon-exclusive to the wider Copilot+ PC universe powered by AMD and Intel. These enhancements are especially relevant for individuals with visual impairments, now able to command “Narrator key + Ctrl + D” to have detailed visual content read aloud.This goes beyond simple alt text, leveraging context-aware AI to extract and vocalize finer shades such as the presence of people, objects, color palettes, text overlays, and even numerical data. On older or non-Copilot+ hardware, the descriptions are more superficial—likely reflecting the limitations of onboard AI accelerators in those systems.
The technical underpinning leverages Microsoft’s broader investments in generative AI models and real-time neural processing, with clear reference to the Windows Copilot infrastructure. Testing has shown that, while performance varies according to underlying hardware, the qualitative leap in accessibility is undeniable. Side-by-side evaluations indicate that the AI engine on x86-based Copilot+ systems delivers parity to ARM devices in most day-to-day image description tasks.
There is, however, a broader privacy conversation around cloud versus local processing of these AI-driven interpretations, particularly when sensitive imagery is involved. Microsoft asserts that on-device NPUs handle processing for Copilot+ PCs, mitigating the risk of data leaving the machine unless explicitly enabled for cloud enrichment. This claim aligns with statements in recent Insider build notes and is supported by independent analysis, although enterprise administrators are well-advised to review privacy policy updates as the feature matures.
3. Lock Screen Widgets: Intelligent Discovery
Microsoft’s steady drive to make the Windows Lock Screen more utilitarian and less of a passive adornment took another step in July. The new “Discover widgets” toggle, present since build 26200.5710 (Dev) and 26120.4741 (Beta), brings intelligent, regularly-updating widget suggestions right to the front of the lock screen. This adjustment is automatic, drawing on user preferences and contextual cues to surface relevant at-a-glance information—from weather updates and news snippets to calendar highlights and traffic alerts.Usability advocates see promise in this approach, envisioning a lock screen that’s both alive and uniquely adaptive to each user. In limited user trials, engagement with suggested widgets increased lock screen interaction by a measurable (albeit regionally variable) margin, according to informal user telemetry shared by Microsoft.
In practice, the feature’s full potential depends heavily on Microsoft’s widget curation algorithms, which will continue to evolve. Early reports note that widget suggestions can sometimes feel generic or redundant, especially for users with highly customized widget dashboards. The trade-off between relevance and information overload is one Microsoft will need to balance carefully.
4. Recent Activity Controls for AI Features
A subtle but critical enhancement in user privacy oversight arrived with updates to the “Text and image generation” controls under the Privacy & security settings. This refined dashboard now includes a “Recent activity” pane detailing which third-party apps have accessed AI models within Windows 11.For security-conscious users and IT administrators, such visibility is increasingly non-negotiable as AI integration accelerates. By surfacing this data, Microsoft essentially shines a light on app behaviors, empowering users to catch suspicious or unwanted usage of generative features like text and image generation.
This transparency push is aligned with broader industry trends post-GDPR and emerging regulatory frameworks on digital sovereignty. Although critics argue that granular, per-process logging is the next logical step, the current “Recent activity” screen is a notable leap towards accountable, auditable AI usage within the consumer OS sphere.
Peer comparison shows that Windows now offers a higher level of AI access self-auditing than macOS or mainstream Linux distributions, where such activity tracing is often buried behind more technical user interfaces or relies on third-party tools.
5. AI Agent in the Settings App: Natural Language, Real Answers
One of the more ambitious, Copilot-branded features to hit preview in July is the new AI agent built into the Windows Settings app. Heralded as a change that will roll formally into the Stable Channel with the upcoming security update in August, this capability lets users query settings, resolve problems, and apply adjustments using plain language—without navigating dense menus or arcane system paths.Previously tested for select Copilot+ environments, the AI agent is now out for a broader cohort of AMD and Intel Copilot+ devices. For instance, typing “Make my system more private” or “Turn on dark mode at sunset” prompts relevant settings or workflows, with the agent able to auto-resolve or guide the user interactively.
This builds on several years of Microsoft investment in natural language processing for productivity (think PowerShell Copilot, Office natural queries). Direct competition exists, from Apple’s growing integration of conversational controls in System Preferences to Google Assistant’s cross-platform reach, but early Insider feedback points to Windows’ agent excelling in multi-step, compound system tasks.
Still, potential gotchas remain. Reports from the field suggest occasional vagueness or refusal to commit action, especially if the underlying setting is multi-faceted or locked by policy within managed environments. There is also an open question about on-device versus cloud execution—the AI agent primarily leverages on-device inferencing via NPU but may still reach out to Microsoft’s servers for complex requests, unless users opt out.
6. Combined Second Chance Out-of-Box Experience (SCOOBE) Recommendations
Windows’ onboarding experience—famously known as OOBE (Out-of-Box Experience)—has long been a point of both pride and criticism. The latest tweak compresses what was once a sprawling set of post-update recommendations and offers after a restart into a single consolidated page. Rebranded as “Second Chance OOBE” (SCOOBE) in July’s previews, the new screen nudges users to enable features like cloud backups or, notably, to opt in to using the Microsoft Edge browser.By aggregating these “second chance” prompts, Microsoft seems intent on reducing user friction and streamlining the persuasion process. The downside? The overt marketing push for Edge, coupled with repetitive offers, sparks renewed debate over where helpful guidance ends and invasive advertising begins. Savvy users will appreciate the cleaner interface; skeptics may bristle at what can feel like a coerced conversion funnel.
Strikingly, this mirrored approach is seen in other operating system ecosystems (e.g., Apple’s iOS prompt to enable iCloud or privacy settings after updating), but its effectiveness in Windows depends heavily on user tolerance for nudges versus perceived pressure.
7. Windows Search: Grid Image Results
Microsoft continues to enhance the core search experience, with July’s Insider builds debuting grid-based image search results in the main Windows Search pane. When users initiate image-related queries, relevant thumbnails now populate in neat, glanceable grids—speeding up the identification and selection process.This small but potent polish applies modern UX tenets, borrowing from visual discovery patterns pioneered by Google and Apple’s Spotlight search. The left-hand image grid accelerates workflows that would otherwise send users into File Explorer–intensive treasure hunts.
According to usability studies and early Insider feedback, this feature is particularly valued by creative professionals, educators, and social media managers who often work with large local media libraries. Potential caveats include thumbnail resolution trade-offs and rare but possible privacy concerns if search indexes protected or system-only folders.
8. Notification Center with Calendar and Clock: Now Available on Secondary Monitors
A frequent complaint among power users with multi-monitor setups has finally been addressed: the Notification Center, along with its embedded calendar and full clock (down to the seconds), is now accessible on every connected monitor. Users need only click the time and date in the System Tray of any display to summon the flyout interface.This fix dramatically enhances daily usability for professionals, developers, gamers, and anyone running dual or triple displays, creating a more unified and less frustrating Windows experience. Previously, secondary monitors were left with static time displays, leaving the Notification Center chained to the primary display. The improvement also brings Windows 11 into tighter feature parity with Linux desktop environments like GNOME and KDE, which have long supported such multi-monitor UI extensions.
9. Task Manager: Standard and Legacy Processor Metrics in Harmony
Rounding out the July updates is a fundamental, if somewhat technical, redesign of how processor usage metrics are displayed in Task Manager. The visual refresh returns industry-standard metrics to the ‘Processes,’ ‘Performance,’ and ‘Users’ pages, following feedback that past updates made CPU monitoring less transparent for power users and IT admins.Users can also, via the Details page and context menu, add a “CPU utility” column—bringing back legacy, granular processor metrics for those who prefer the old readouts. This hybrid model is welcome news for developers, system optimizers, and hardware enthusiasts who want both simplicity and depth in their performance monitoring tools.
Competing operating systems such as macOS and Ubuntu have their own conventions, but Windows’ dual-mode approach is arguably the most user-configurable, bridging the needs of novices and veterans alike.
The Bigger Picture: What These Changes Mean for Users
Collectively, these features paint a picture of Windows as a platform doubling down on three interwoven strategies: AI-centric user empowerment, deeper accessibility, and responsive transparency. Microsoft’s dedication to real-time AI features—now scaled across all processor architectures and underpinned by Copilot+ branding—marks the beginning of a new, hardware-accelerated OS paradigm.The improvements in accessibility, widget interaction, natural-language controls, and on-device privacy audits establish a model for responsible, useful AI. At the same time, careful attention is required as Microsoft pushes cloud-linked features and targeted recommendations (as seen in the combined SCOOBE), lest this tip from helpfulness into overreach.
Noteworthy Strengths
- Widespread AI accessibility: Many Copilot+ features, especially image description and AI agent assistance, are now available beyond Snapdragon-equipped devices—unlocking inclusive and equitable functionality for a much broader user base.
- User-oriented transparency: From privacy dashboards to on-device processing of AI tasks, Microsoft is surfacing more internal OS behaviors than ever before.
- Workflow acceleration: Grid-based image search, multi-monitor notifications, and legacy metrics coexistence each cut friction in daily use, accelerating common tasks for novices and veterans.
- Accessibility at the core: Advanced Narrator features and Click to Do improvements are not just bolted on; they are fundamental usability upgrades, fulfilling both regulatory compliance and real societal need.
Potential Risks and Open Questions
- Consistency across hardware: Early testers report uneven performance and occasional glitches in AI-powered features outside Snapdragon hardware, hinting at inevitable growing pains as the Copilot+ umbrella spreads.
- Information overload: Widget and notification creep, if not managed, could overwhelm users, especially as lock screen widgets and SCOOBE prompts proliferate.
- Privacy nuances: Despite strides in keeping processing local, certain features may still fall back to cloud processing for “difficult” tasks. Clear user opt-outs and policy communication will be critical.
- User autonomy: The consolidation of recommendations (particularly those nudging Edge adoption) walks a fine line. Too much persuasion risks alienating the informed user base that makes up much of the Insider audience.
Looking Ahead: The Windows Evolution Continues
July’s preview cycle should hearten anyone invested in Windows’ future. Microsoft is iterating quickly, driven by both competitive pressures and deeper engagement with its enthusiast community. The threefold focus—more inclusive AI, a modernized work surface, and new privacy guarantees—sets a compelling, if challenging, course for the months ahead.As the 24H2 and 25H2 updates near release, seasoned observers will be watching for refinements in feature reliability and further democratization of previously exclusive Copilot+ capabilities. For power users, admins, and accessibility champions, Windows 11 is in the midst of its most significant transformation in a decade—one that feels both grounded in practical benefit and alive with technical ambition.
Source: Windows Central 9 new features that arrived on the Windows 11 Insider Program during the second half of July 2025
