BloomingTalk's AI Idol Voices Draw Renewed Fan Backlash
BloomingTalk, which officially launched in August 2025, drew renewed criticism in early August 2026 over paid AI chats and voice calls modeled on A2O MAY members. Koreaboo documented fan concerns about authenticity and emotional reliance, while the app's own listing identifies the conversations as AI and describes Secret Chat as private between the user and the AI artist.
BloomingTalk, a fan platform that combines direct artist messages with AI-generated chats and voice calls, drew renewed criticism in early August 2026. Koreaboo documented reactions posted on August 2 and 3 questioning whether an always-available synthetic version of an idol should be marketed as an authentic connection.
The timing matters. Music Business Worldwide and Digital Music News reported in July 2025 that BloomingTalk was preparing to launch the following month. An A2O MAY post embedded in the current Koreaboo report dates the official launch announcement to August 6, 2025. The present story is therefore about renewed scrutiny of an existing product, not a new 2026 launch.
How the service works
BloomingTalk separates direct artist communication from automated interaction. Koreaboo describes "Human Talk" as messages sent by artists and "AI Talk" as responses generated in an artist's conversational style, including voice calls that recreate the participating artist's voice. The Apple App Store listing similarly presents the product as AI conversation and says Secret Chat is a private thread visible only to the user and the AI artist.
Koreaboo reports that an AI Talk pass for an A2O MAY member costs KRW 6,800 and that additional message credits are sold after the included allocation is used. Those prices are reported figures rather than a permanent rate: BloomingTalk's user agreement says paid-service pricing may change.
Earlier 2025 reporting said the service was designed to support multiple languages and that participating artists had agreed to fixed contract terms. Music Business Worldwide also reported that the company was developing persona rules restricting sexual material, inappropriate requests, and offensive language. The retrieved public material does not disclose the model architecture, voice-synthesis provider, enforcement performance, or detailed retention controls.
The product-design question
The current backlash is not evidence that BloomingTalk hides its automated mode; the product distinguishes Human Talk from AI Talk. The sharper concern is how a paid, always-available replica is framed to fans and whether the interface makes the boundary between artist-authored and model-generated interaction unmistakable throughout the experience.
For teams building synthetic-persona products, the case highlights four controls that need explicit treatment: performer consent, persistent AI labeling, limits on emotionally manipulative engagement, and clear handling of private conversations. The public evidence supports scrutiny of those design choices, but it does not establish misuse of an artist's voice or a breach of the stated participation agreements.
Key Points
- 1BloomingTalk launched in August 2025; the August 2026 event is renewed fan criticism, not a new product launch.
- 2The service separates artist-authored Human Talk from paid AI chats and voice calls modeled on participating A2O MAY members.
- 3The available evidence supports questions about labeling, consent scope, privacy, and engagement design, but does not establish unauthorized voice use.
Scoring Rationale
The renewed criticism concerns a real consumer deployment of paid AI personas and voice interaction. It is useful to product teams working on consent, disclosure, conversational boundaries, and engagement design, while the platform's immediate reach remains comparatively narrow.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
