Babbel v22.5.0 uses voice data for AI training without opt-out, sparking outrage in July 2026

The official release notes for version 22.5.0 provided no details about the data policy change. The changelog on the Apple App Store simply stated, "We've been working hard on the app to improve your learning experience." This lack of transparency hid a critical change that directly contradicted many users' privacy expectations, setting the stage for the negative reception.^[2]

The central issue for users is the non-consensual use of their voice recordings. A one-star review for version 22.5.0 captures the sentiment: "I bought the year long subscriptionGlossarySubscriptionA recurring in-app purchase that charges the us^[3]er on a fixed cadence (weekly, monthly, annual) until they cancel. only to find out AFTER that they use your voice to train their AI. No opt out option Immediately cancelled." This user further stated, "I am not a vessel from which to scrape my data and monetize every last scrap of my humanity and privacy."

The policy change has led to immediate subscription cancellations, impacting a user base of over 227,000 monthly downloads on US iOS. Other complaints surfaced after the update, including one user on v22.5.0 noting, "They ask every day for a rating. Every day. I’ve submitted about 100 ratings. Enough is enough."

The sentiment shift is stark compared to reviews on previous versions. Before the update, feedback often focused on feature requests and performance, such as one four-star review on v22.4.1 that noted, "the speech recognition is having issues and not able to hear to my voice." Now, the conversation is dominated by privacy concerns.

Discussions outside the App Store confirm the user reaction. A thread on Reddit titled "Babbel is using your voice to train their AI with no opt out option" mirrors the core complaint, with users expressing intent to cancel their subscriptions. Archived.

Babbel's decision was likely a strategic move to improve its speech recognition technology, a key feature for language apps. By gathering more training data, the company could enhance lesson effectiveness and gain a competitive edge. However, the implementation failed to account for user sensitivity to data privacy. The critical error was not the use of data itself, but the failure to provide a simple, accessible opt-out, a standard practice for privacy-conscious software.

Babbel's path forward requires addressing the privacy backlash directly. The drop in ratings and public cancellation of subscriptions signal a clear financial and reputational risk. To repair user trust, the publisher will likely need to issue a public statement and ship a subsequent update that introduces transparent data controls, including a clear opt-out for AI training. Without such action, the negative sentiment is likely to persist, potentially harming new user acquisition and increasing churnGlossaryChurnThe rate at which users stop using an app (general churn) or cancel a subscription (subscription churn) over a time window — retention's mathematical complement..