Google tracks how people use its Gemini AI assistant with far more precision than most users probably suspect. Court documents filed in the federal antitrust case against the company reveal that Alphabet’s board of directors received a dedicated Gemini briefing in October 2024, complete with monthly and daily active-user metrics. The presentation, now partially visible to the public through the U.S. Department of Justice, offers the clearest signal yet that Google has built internal tooling to monitor AI consumption closely. Based on the existence of board-level MAU and DAU tracking, it is reasonable to infer that Google possesses the infrastructure to meter AI usage at scale, though the documents do not explicitly describe such a system in consumer-facing terms. That inference raises a pointed question: are consumer-facing usage limits next?
What the court documents actually show
The key exhibit is Trial Exhibit PXR0226, a Google slide deck titled “Gemini App Alphabet Board Update,” dated October 2024. The DOJ posted a redacted PDF of the presentation as part of its public remedies exhibit repository in the case styled U.S. and Plaintiff States v. Google LLC. Even with heavy redactions, the visible portions confirm that Google tracked Gemini’s monthly active users (MAU) and daily active users (DAU) and presented those growth figures directly to Alphabet’s directors.
That detail matters more than it might seem at first glance. Companies reserve board-level product updates for offerings they consider strategically critical. By giving Gemini its own dedicated briefing, Google placed the AI assistant in the same tier of executive attention as Search and YouTube. The existence of board-grade usage dashboards also suggests, by reasonable inference, that Google operates internal tooling capable of measuring AI consumption in near real time, with the rigor directors expect before making resource-allocation decisions. This inference follows from the nature of board reporting but is not explicitly stated in the visible portions of the deck.
The same exhibit index lists Samsung-Gemini distribution agreements, confirming that Google negotiated device-level placement for Gemini on Samsung hardware, mirroring the preinstallation strategy it has long used for Search on Android phones. The specific contract terms remain sealed, but their appearance in a high-stakes antitrust proceeding signals that hardware-integrated distribution is central to Gemini’s growth plan, not an afterthought.
Why usage metering matters right now
Running large language models is expensive. Every query a user sends to Gemini consumes compute cycles on specialized AI chips inside Google’s data centers, and those costs scale directly with demand. Google is not alone in grappling with this: OpenAI introduced and repeatedly adjusted usage caps for ChatGPT Plus subscribers throughout 2024 and into 2025, and Anthropic’s Claude Pro plan has long carried conversation limits during peak hours. Microsoft’s Copilot similarly throttles certain features for free-tier users.
Google itself has already moved in this direction on the developer side. In early 2025, the company tightened rate limits on its Gemini API tiers, reducing the number of free requests per minute for models like Gemini 1.5 Pro and later adjusting quotas again when Gemini 2.5 Pro launched. Those API-level controls demonstrate that Google has both the technical capability and the institutional willingness to cap AI usage when costs or capacity demand it.
What the board deck adds to this picture is executive-level visibility. Rate limits set by a product team are one thing; MAU and DAU charts reviewed by Alphabet’s directors suggest that usage management is a strategic priority, not just an engineering constraint. When a company’s board is watching consumption curves, decisions about tiered pricing, free-access caps, or usage dashboards often follow, though that outcome is an inference from industry patterns rather than something the documents state outright.
What the documents do not confirm
It is worth being precise about the gap between what these exhibits show and what they imply. The redacted deck does not contain any visible slide describing a consumer-facing usage dashboard. It does not spell out specific MAU or DAU numbers, regional breakdowns, or explicit plans to impose caps on free Gemini access. The Samsung agreements, while confirmed to exist, do not reveal whether Samsung would play any role in enforcing usage limits on its devices or whether the contracts simply govern app preinstallation and promotional placement.
The headline’s reference to a “leak” that “points to a dashboard” deserves a caveat on both counts. First, these documents were not disclosed by a whistleblower or obtained through unauthorized access; they were posted on a .gov domain as part of a formal legal process. Google prepared the deck for its own board, and the DOJ introduced it as evidence during the remedies phase that followed the court’s liability finding. Second, no dashboard for tracking consumer usage limits is described anywhere in the visible exhibits. The connection between the internal metrics confirmed in the deck and a hypothetical consumer-facing dashboard is an inference drawn from industry trends and Google’s own behavior on the developer side, not a claim supported by explicit language in the documents. The news value comes from the gap between what Google intended as a private boardroom briefing and what the public can now see, even in partial form, through standard legal channels.
No Google executive has made a public statement, as of May 2026, about plans to roll out a consumer-facing usage dashboard for Gemini. The DOJ’s exhibit repository is a collection of evidence, not an analytical brief, so the department itself has not drawn a direct line between Gemini’s internal tracking and any proposed remedy involving usage limits.
What Gemini users should watch for
For the roughly 300 million monthly users Google claimed for Gemini-integrated products by late 2024, the practical question is straightforward: will Google start showing you how much AI you have used and how much you have left?
The pattern across the industry suggests the answer is eventually yes, in some form. Every major AI provider has moved toward usage transparency and tiered access as models grow more capable and more expensive to run. Google already operates a usage-tracking interface for Google One storage and maintains detailed quota dashboards for its Cloud API customers. Extending a similar concept to consumer Gemini would be technically straightforward for a company that, as the court documents confirm, already measures consumption at the board-reporting level.
The timing and shape of any such dashboard remain genuinely unknown. Google could introduce soft notifications (“You’ve used 80% of your free queries this month”), hard caps that push users toward Gemini Advanced subscriptions, or something more nuanced tied to model complexity, where lightweight queries stay free but resource-intensive tasks like long-document analysis or image generation count against a quota. All of these approaches have precedents at competing platforms.
Privacy is another open thread. The board deck confirms that usage is tracked and aggregated, but it does not reveal what identifiers are attached to those records, how long they are retained, or how they intersect with Google accounts and advertising profiles. Those details will matter enormously if Google ever surfaces usage data directly to consumers, because a dashboard that shows you your own AI habits also implies a backend that stores them.
The remedies case is still producing new exhibits
The antitrust proceeding that surfaced these documents is far from over. The original complaint was filed in 2020, and the remedies phase, which began after the court found Google liable, has continued to pull in contemporary internal materials. As more exhibits are filed and potentially unredacted, the public picture of Google’s Gemini strategy will sharpen.
For now, the court record confirms something specific and significant: Google built a sophisticated internal apparatus to track Gemini’s user base, reported those metrics to Alphabet’s board, and pursued hardware distribution deals with Samsung to expand the assistant’s reach. From those facts, it is reasonable to infer that Google possesses the infrastructure to meter AI usage at scale, but that inference should not be confused with proof. The documents do not confirm that a consumer-facing usage dashboard is imminent or that specific caps on free access have been approved. The internal measurement infrastructure is documented. What Google plans to do with it, publicly, remains an open question worth watching closely as the remedies hearings continue through 2026.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
