Category: Technology | Published: 2026-06-09
When you buy a physical product, there is a reasonable expectation that the things it does when you take it home are the things it will continue to do. That expectation is increasingly complicated when the product is a connected device and the things it does depend on software that its manufacturer can update, adjust, or restrict at any time.
Meta has just demonstrated this with its AI glasses. A feature built into the hardware, available to all users at launch, has now been capped at a monthly usage limit for anyone who does not subscribe to a paid plan. The move has attracted criticism, raised questions about how the cap is even justified technically, and opened a broader conversation about what buying a piece of connected hardware actually means in 2026.
What Changed and What It Now Costs
The feature in question is called Conversation Focus. It is built into Meta's Ray-Ban AI glasses and uses the glasses' microphones and speakers to amplify the voice of the person directly in front of the wearer, while reducing the ambient noise around them. In a noisy bar, a busy office, a crowded conference, or any environment where background noise makes conversation difficult, it makes the person you are talking to significantly easier to hear.
This feature is now free for three hours per month. Anyone who wants more than that needs to subscribe to Meta One Premium, which costs $19.99 a month and raises the allowance to fifteen hours per month.
To put those numbers in everyday terms: the free allowance works out to roughly six minutes per day if spread evenly across a month. The paid tier comes to approximately thirty minutes per day. Unused hours do not carry over into the following month under either plan, so there is no way to save time from a quieter period for a busier one.
Other AI glasses features remain available without a subscription. Meta is clear on this point, stating that there is no subscription required to use AI glasses, and that users will continue to have access to AI glasses features without a Meta One subscription. Conversation Focus is the most prominent feature to have been placed behind a usage cap so far.
The Question Nobody Has Answered
One of the more pointed criticisms of Meta's decision concerns the technical nature of Conversation Focus itself. Unlike features that involve sending data to remote servers for processing, generating AI responses in the cloud, or accessing Meta's broader infrastructure, Conversation Focus appears to operate entirely on the glasses using their onboard hardware. The microphones capture audio. The speakers deliver it. The processing happens locally.
If that is the case, there is no obvious server-side cost to Meta when someone uses the feature for three hours versus fifteen. The monthly cap would not be limiting server load or managing infrastructure costs. It would be a commercial decision to create a usage ceiling on something that works offline, packaged with language about usage limits rather than paywalled features. The research notes that by calling it a usage cap rather than a paywall, Meta creates some legal and reputational distance from a bait-and-switch framing. Meta has not explained whether Conversation Focus draws on its infrastructure in any way.
Who This Affects Most
Meta's position is that the vast majority of users will not reach the three-hour free monthly limit. For casual users who activate Conversation Focus occasionally at a loud event, that is probably true.
The problem is that the people most likely to use Conversation Focus regularly are precisely the ones for whom it provides the most genuine value: people who find conversations difficult in noisy environments, those who rely on the feature during business meetings held in open-plan offices or busy restaurants, and anyone whose working life involves regular face-to-face communication in places where background noise is a constant. For those users, six minutes a day is not generous.
The feature is not marketed as a medical device or a hearing aid, but the practical overlap with that use case is obvious. Placing a usage cap on something that helps people hear conversations more clearly has attracted a different quality of criticism than a limit on, say, a creative AI feature for generating images.
The Broader Pattern: From Hardware to Hardware-Plus-Subscription
Meta's AI glasses decision is a visible example of a transition that has been happening across connected technology for several years. The traditional model was simple: you buy the product, you get the features. The emerging model is more complicated: you buy the product, you get a baseline of features, and additional or unrestricted access to capabilities built into the hardware requires an ongoing subscription.
This pattern is already established in business software, where perpetual licences have largely given way to monthly and annual subscriptions. It has spread into cars, where manufacturers including Tesla, BMW, and others have offered features like heated seats or advanced driver assistance systems as subscription upgrades to hardware that is physically present in the vehicle. It appears in home security cameras, smart doorbells, robot vacuum cleaners, and domestic appliances.
Smart AI glasses are particularly well suited to this model because a significant proportion of their value comes from software and AI services rather than the physical frames. New capabilities can be introduced remotely. Existing capabilities can be restructured into tiers. The hardware becomes a platform for ongoing commercial relationships rather than a finished product.
The True Cost of Ownership Problem
For consumers and businesses evaluating AI glasses and similar connected devices, Meta's decision raises a practical question that is becoming more important: how do you calculate the true cost of ownership when the feature set is not fixed at the point of purchase?
A device that costs a certain amount upfront looks different in a total cost calculation once a monthly subscription is added for features that were previously included. If further features move behind paid tiers over time, the ongoing cost increases again. The upfront price becomes less representative of what you will actually spend.
This is not necessarily a reason to avoid connected hardware. But it is a reason to understand the manufacturer's commercial model before buying. What is included permanently? What has usage limits? What might be restructured into a paid tier in a future update? Those questions are worth asking before a purchase rather than discovering the answers afterwards.
What Could Come Next for AI Glasses
Meta has been explicit that Meta One is designed to provide greater access to AI features across its range of products, not just AI glasses. The subscription currently sits alongside expanded access to features on Instagram, Facebook, and WhatsApp as well as the glasses-specific capabilities.
As Meta's AI glasses become more capable, the number of features that could be structured into paid tiers will grow. Manufacturers who have committed to a subscription model have a commercial incentive to continue expanding it. Whether that means more features moving behind usage caps, higher tiers with additional capabilities, or both will become clearer as the product evolves.
What This Means for Businesses
For businesses thinking about AI tools and connected devices, the subscription question is increasingly a budget planning question rather than just a consumer concern. Software-as-a-service pricing is already standard in business technology. The extension of subscription models into hardware-embedded AI features means that total cost of ownership calculations need to account for ongoing fees that may not be immediately obvious at the point of purchase.
Building AI tools into your workflows in a way that is sustainable and genuinely cost-effective requires understanding not just what a tool does today but what it is likely to cost over time and what happens if pricing changes after adoption. Our AI Consultancy page covers how we help businesses evaluate and adopt AI tools with a clear picture of long-term value rather than short-term enthusiasm.