Meta pauses smart glasses feature paywall after backlash

Meta ↗ 1.44% has paused its plan to cap the Conversation Focus accessibility feature on its smart glasses at three free hours per month, retreating from a pricing structure that drew sharp criticism from users and accessibility advocates over the past several weeks.

The Reversal

The Verge reported on Thursday that Meta walked back the rate-limiting plan after sustained backlash from the glasses' user community. The feature, which isolates and amplifies the voice of whoever a wearer is facing in a noisy environment, was set to require a $19.99-per-month Meta One Premium subscription for expanded use of up to 15 hours monthly. Meta spokesperson Tyler Yee confirmed the pause, stating that Conversation Focus "will continue to be offered for free via our Early Access Program for initial testers as we strive to develop an improved solution". Yee indicated that some premium features will still become subscription-based over time.engadget+5

Why the Backlash Was So Sharp

Meta quietly introduced the rate limits on its help page around June 30, as first reported by The Verge. The cap worked out to roughly six minutes of daily use on the free tier. What made the outcry particularly loud was the technical nature of the feature itself: Conversation Focus runs entirely on-device using beamforming microphones and spatial audio processing, requiring no internet connection and generating no server costs for Meta.androidauthority+3

That undercut the typical justification for AI subscription models, where cloud inference carries real per-query expenses. For hard-of-hearing users, the feature functions as assistive technology rather than a novelty, making the paywall feel especially punitive. Even paying subscribers would have been locked out after 15 hours, with no option to roll over unused time to the following month.aiweekly+3

What Comes Next

It remains unclear whether Meta has scrapped the cap entirely, softened the limits, or simply delayed the rollout while reworking its messaging. The episode offers a cautionary signal for hardware makers pursuing recurring revenue from on-device AI features: users appear willing to pay for cloud-backed services but resist subscriptions for capabilities that run on silicon they already own.aiweekly