Enterprise AI Governance

Nexthink AI Activation Hub

AI Activation Hub powered by AI Drive is Nexthink's platform for managing enterprise AI from discovery to ROI. It finds every AI tool running across the fleet, including unsanctioned shadow AI. It enforces governance policies, delivers in-the-moment employee guidance, and gives leadership the usage and ROI data they need to make decisions on AI investment. Nexthink positions it as a way to "measure, drive and improve AI adoption." Our practice applies it to observation and governance.

Last reviewed November 2026

Overview

The Enterprise AI Management Layer

AI Activation Hub sits at the organizational layer of Nexthink's AI capabilities. It does not automate resolution for employees (that is Spark) or provide an AI cockpit for IT analysts (that is Workspace). It answers a different set of questions: what AI tools are actually in use across the enterprise, which are sanctioned, who is using them deeply versus not at all, what is the experience quality during use, and how do you translate that into governance decisions and ROI reporting?

It uses the same Nexthink Collector infrastructure and endpoint telemetry as the rest of the platform, extended with browser-level monitoring for AI web tools and API connectors for SaaS AI services. Because it operates at the endpoint rather than at the network perimeter, it captures usage patterns that proxy-based tools miss, including tool-switching behavior, session duration, and the device conditions that affect whether an AI tool delivers its intended productivity gain.

Framework

The Five-Pillar Framework

AI Activation Hub structures enterprise AI management across five disciplines. Each builds on the one before it: you cannot govern what you have not discovered, and you cannot measure ROI without first assessing actual usage depth.

Discover

Find every AI tool in use across the enterprise, including shadow AI that was never sanctioned or inventoried. Browser extension monitoring and endpoint telemetry surface tools that employees have started using independently, outside IT's awareness.

Assess

Measure usage depth, usage frequency, and experience quality for every AI tool in the inventory. Cohort analysis shows which departments are using Copilot actively and which use it in name only. Device and network telemetry explains why experience quality varies.

Govern

Enforce AI usage policies across the fleet. Flag unsanctioned tools that violate data handling policies, identify compliance risks from specific AI applications, and give IT and security teams an accurate, continuously updated inventory to act on.

Enable

Deliver in-the-moment guidance to employees at the point of AI tool use through Nexthink Guides. Contextual nudges, feature prompts, and usage tips are surfaced directly on the employee's device based on actual behavior, not scheduled training programs.

Measure

Report AI ROI to leadership with data procurement and finance can act on. Usage rates by team, trend over time, correlation with digital experience quality, and identification of IT-solvable blockers give organizations the evidence base for AI investment decisions.

Key Capability

Shadow AI Discovery

Shadow AI is the category of AI tools employees are using that IT never approved or even inventoried. It includes consumer-grade tools like ChatGPT accessed via browser, AI coding assistants installed as browser extensions, and vertical AI SaaS products brought in by individual teams outside the standard procurement process. Most enterprises significantly undercount their AI exposure when they rely only on software asset management records or network-level monitoring.

Browser extension monitoring

AI Activation Hub uses browser-level telemetry to detect AI tools operating as extensions or web applications. This captures usage that never appears in network logs because it operates over standard HTTPS traffic indistinguishable from ordinary browsing. ChatGPT, Grammarly with AI, Jasper, Perplexity, and similar tools are visible at this layer.

SaaS connectors

API connectors to major AI platforms supplement endpoint telemetry with usage data pulled directly from the source. This gives a usage record that can be correlated with endpoint conditions: an employee using Copilot heavily while on a slow VPN connection shows up as an experience quality issue, not a Copilot usage success.

Risk classification

Discovered AI tools are classified by data handling risk, based on whether they process data locally, transmit to third-party servers, or have privacy policies that conflict with enterprise data governance requirements. IT and security teams get an actionable risk inventory, not just a raw list of tool names.

Policy enforcement

Shadow AI findings feed directly into governance workflows. IT can configure policies that trigger automated alerts, restrict access to specific tools, or surface warnings to employees using non-compliant AI applications. Enforcement is proportional: notification for low-risk tools, restriction for high-risk ones.

Enable Pillar

Nexthink Guides: Contextual AI Enablement

Nexthink Guides is the delivery mechanism for the Enable pillar. It surfaces contextual, in-the-moment guidance to employees at the point of AI tool use, based on what the platform observes about their actual behavior rather than a fixed training schedule.

Usage nudges

Employees who have access to Copilot but have not used it in two weeks receive a contextual prompt the next time they open Word or Teams. The prompt is triggered by observed behavior, not a calendar reminder, making it more likely to land at a relevant moment.

Usage tips at the point of use

Employees actively using an AI tool can receive contextual tips about features they have not yet used, based on their actual usage pattern. A Copilot user who only uses it for email drafting might receive a prompt about meeting summarization when they join a Teams call.

Policy notifications

When an employee attempts to use a shadow AI tool that has been flagged, Guides delivers a notification that explains the policy, offers an approved alternative, and provides a one-click path to request a review. This is materially different from a blocked page: it turns a restriction into a guidance moment.

Segment-specific campaigns

Guides campaigns can be scoped to specific employee cohorts based on department, role, or observed AI tool usage patterns. Sales teams get Copilot guidance relevant to CRM workflows; developers get guidance specific to GitHub Copilot. Broad awareness campaigns are replaced by targeted behavioral nudges.

Clarification

How AI Activation Hub Differs from Nexthink's Other AI Capabilities

Nexthink has several AI-related capabilities that serve different audiences and answer different questions. The distinction matters when scoping an implementation or positioning the value to stakeholders.

Capability	Audience	What it does	What it does not do
Spark	Employees	Autonomous personal IT agent: diagnoses and resolves device issues without a ticket, escalates when human intervention is needed	Does not govern AI tool usage or report on enterprise AI usage
Workspace	IT analysts	Conversational AI cockpit for investigations: natural language to NQL, proactive fleet insights, multi-step agentic reasoning	Does not discover shadow AI or measure Copilot usage across the organization
AI Insights	IT analysts	Pre-built AI-generated insights surfaced in the Nexthink platform: anomaly detection, experience score drivers, trend analysis	Does not enable interactive investigation or conversational follow-up (that is Workspace)
AI Assist	IT analysts	Natural language interface for platform navigation, NQL generation, and Engage campaign drafting within the Nexthink console	Does not run multi-step investigations autonomously (the distinction from full Workspace)
AI Activation Hub	IT, Security, Leadership	Enterprise AI governance platform: shadow AI discovery, usage observation, policy enforcement, employee enablement, and ROI reporting	Does not automate IT resolution (Spark) or provide an investigation cockpit (Workspace)

Measure Pillar

Reporting AI ROI to Leadership

The Measure pillar is what closes the loop between AI investment and business outcomes. Most AI ROI reporting fails because the data is self-reported (surveys) or incomplete (license activation counts). AI Activation Hub provides behavioral data from the endpoint layer: actual usage, not logins.

Usage by department and cohort

Usage rates are segmented by department, team, role, and geography. Finance and procurement can see that Sales has 78% active Copilot usage while HR is at 23%, and that the HR gap correlates with a known device age issue in that group, not low motivation.

Trend over time

Usage trend data shows whether it is growing, plateauing, or declining after initial rollout. This distinguishes between tools that employees genuinely find valuable and those that were used once after a mandatory training session and then abandoned.

Experience quality correlation

Usage depth is correlated with the underlying device and network experience during AI tool sessions. Low usage in a specific team may not be a change management problem. It may be that the team's aging devices cannot run the local AI inference that Copilot requires, resulting in a poor experience that discourages repeat use.

IT-solvable blockers

AI Activation Hub surfaces usage blockers that are within IT's control to fix: hardware that cannot support AI tool requirements, network configurations that add latency to AI inference calls, software conflicts that cause AI tools to crash. These are presented as actionable items, not just data points.

Release Updates

AI Activation Hub: What's New in 2026.5 to 2026.8

In the product, the AI Activation Hub capabilities appear as AI Tools, powered by AI Drive. Recent releases strengthened the governance side considerably.

2026.5 (June 2026): AI Tools Governance detects AI-related domains and applications in use and highlights tools outside the sanctioned list, with built-in workflows to review, allow, or flag them. This is the core shadow AI control.
2026.6 (July 2026): Cascading organizational filters make it easier to break AI usage down by division, department, and team.
2026.7 (August 2026): The discovered tools list now shows how many employees use each tool and when it was first seen, which helps prioritize reviews and spot new tools quickly. Policies (Recommended, Allowed, or Prohibited) can be assigned during configuration in one step, and the built-in AI Tools employee campaign can be previewed before publishing.
2026.8 (September 2026): Organization-wide goals (Nexthink calls them "AI adoption goals") can be set with a metric, target, and scope, then tracked over time and broken down by department, region, or role. The weekly AI engaged time chart now shows up to 52 weeks of history, up from 13. Some capabilities may require additional licensing.

Implementation Guidance

What You Need Before AI Activation Hub Delivers Value

AI Activation Hub relies on the same Collector infrastructure as the rest of the Nexthink platform. The prerequisites are largely the same, with a few additions specific to AI tool monitoring.

High Collector coverage

Shadow AI discovery is only as complete as Collector coverage. A fleet with 70% Collector deployment means 30% of AI tool usage is invisible. For organizations where the shadow AI discovery finding is the primary driver of value, coverage gaps directly limit that value. Target 95%+ before relying on the inventory for governance decisions.

Browser extension deployment

Shadow AI discovery via browser monitoring requires the Nexthink browser extension deployed to managed devices. This is a separate deployment step from the Collector. Organizations that rely solely on endpoint telemetry without the browser extension will miss web-based AI tools that account for a large share of shadow AI usage in most enterprises.

A defined AI governance policy

AI Activation Hub enforces policies, but it does not write them. Before the Govern pillar can operate, IT and security leadership need to define which AI tools are sanctioned, which are prohibited, and which fall into a review queue. Without a policy, the inventory data has nowhere to go: enforcement campaigns cannot be configured against an undefined standard.

Connector configuration for licensed tools

Usage data from major AI platforms like Microsoft 365 Copilot requires API connector configuration between Nexthink and the relevant SaaS platform. This involves admin permissions in both systems and may require coordination with your Microsoft or Google admin. Verify connector availability and setup requirements with Nexthink before scoping the integration.

Engage and Guides configured

The Enable pillar depends on Nexthink Engage and Guides being active and tested. Organizations that have not yet built out their Engage campaign library, or that have not deployed Guides to end users, will not be able to deliver the contextual nudges that make the Enable pillar work. These are not one-time setup steps: they require ongoing content maintenance as AI tools and policies evolve.

Baseline data before measuring trends

ROI trend reporting requires historical data to establish a baseline. A deployment with two weeks of data can show current usage rates but cannot show whether usage is growing or declining. Plan for a 60-90 day baseline period before presenting trend data to leadership as evidence for AI investment decisions.