Most enterprise AI rollouts follow a familiar pattern. Purchase a license, make an announcement, and declare adoption a success.
The problem is that nobody checks. Weeks later, a significant share of the workforce has logged into Gemini once or twice, clicked around, and gone back to doing things the old way.
Google’s new reporting update, rolled out on February 17, does not solve that problem, but it does make it visible. For the first time, Workspace admins can see exactly who is using Gemini, which features they are using, and who has already hit their monthly usage limit.
What the new admin reports actually show
The reports live inside the Admin console under Generative AI and work at two levels.
At the organisational level, admins can view total active Gemini users, the percentage of assigned licences actually in use, daily usage trends, adoption broken down by app, and interaction counts across Gmail, Docs, Slides, and Sheets. Everything covers the previous 28 days and can be filtered by organisational unit or group.
At the user level, each person gets a usage classification. The top 10% of users with 20 or more interactions are marked ‘high’. Five or more interactions is ‘medium’. Below that is ‘low’, and then ‘zero’. Admins can also see how many days each user was active, and usage broken down by individual app.
Beyond the Admin console, usage and log reports can now be exported to BigQuery for deeper analysis, giving security and analytics teams access to richer data for monitoring adoption trends and identifying training needs at scale.
Why this landed when it did
Google revised usage limits on a subset of advanced AI features on February 3, meaning organisations are now subject to hard caps on things like video generation in Vids, audio overviews, and Workspace Studio automation flows.
The caps vary significantly by plan. Workspace Studio flow executions, for example, are capped at 400 per month on Business Standard and Plus plans, rising to 2,000 on AI Expanded Access and 10,000 on AI Ultra. Audio overviews for PDFs go from 20 per day on standard plans to 200 per day at the highest tier. Further limits on video and image generation features came into force from March 1, 2026.
When users hit those limits, they lose access to those features until the next monthly reset. The new reporting lets admins see who is approaching that point, and plan licence upgrades before disruption occurs.
The adoption gap nobody wants to talk about
The threshold reporting is practically useful. But the more revealing data in these reports will likely be the zero-usage column.
Google’s own research makes the disconnect explicit. Derek Snyder, Director of Product Marketing at Google Workspace, notes that “executive optimism outpaces reality on the ground,” with executives significantly more likely than employees to say AI has had a meaningful positive impact on their organisation. The same research found that only 3% of organisations are highly transformed with AI, while 72% are still in early stages.
The numbers behind that gap are striking. While 61% of employees report using AI daily, only one-in-three feel prepared to adapt to AI-driven changes. Executives are 15% more likely than employees to say AI has had a significant positive impact: a gap that suggests leaders and workers are not experiencing the same rollout.
Visibility is only the start
The harder question is what organisations do with what they find.
Low adoption numbers carry different explanations depending on the team. Some workers may not have relevant use cases. Others may have tried Gemini and found the outputs unreliable for their specific tasks. Others still may not know it exists, or may be working around IT governance concerns without clear guidance.
According to Nirit Cohen, Future of Work Strategist:
“We need to communicate in a way that brings people along because some people are scared. They need to know that if AI helps them become 50% more productive, it won’t simply result in 50% more work.”
Knowing that a team has low Gemini usage does not tell you which of those explanations is true. It gives you a starting point for a conversation, but it requires HR, L&D, and operations teams to follow through with structured enablement rather than another all-staff email about AI.
Dr. Terri Horton, AI Strategy and Workforce Transformation Consultant, makes the point that the approach to that conversation matters as much as having it: “Employees are less likely to fear AI when they understand the organisation’s overarching AI vision and strategy, are upskilled in human and AI collaboration, and have a voice in implementation and responsible use practices.”
AI productivity is now a managed resource, not a licence
Perhaps the most important shift these reports represent is conceptual. AI productivity inside Workspace is no longer unlimited. It is tiered, metered, and capped.
As Google CEO Sundar Pichai said recently at an internal all-hands meeting:
“In this AI moment, I think we have to accomplish more by taking advantage of this transition to drive higher productivity. There will be companies which will become more efficient through this moment in terms of employee productivity, which is why I think it’s important to focus on that.”
For enterprise customers, the implication is direct. Deploying Gemini licences is not an AI strategy. Tracking who uses them, which features drive value, and where enablement is stalling — and adjusting accordingly — is closer to one.
Speaking at the India AI Impact Summit on February 19, Pichai reinforced the workforce dimension of this: “AI will undoubtedly reshape the workforce — automating some roles, evolving others, and creating entirely new careers. Training is crucial.”
Read more:
Read the full article here