Zendesk’s product team is constantly shipping updates, and it can be a full-time job just keeping up. This week, we’re looking at a small but important change for anyone using AI on mobile. Get our expert breakdown of the new AI Label Redesign for Mobile SDKs.
Less Clutter, More Clarity: Our Take on Zendesk’s New Mobile AI Labels
What We Think:
This is a smart, customer-driven update that directly addresses real-world feedback. By redesigning the AI label and making the per-message disclaimer optional, Zendesk is striking a much-needed balance between regulatory transparency and a clean user experience. Our verdict: This is a fantastic quality-of-life improvement, especially for businesses using SMS and social messaging, where every character (and cent) counts. It’s a subtle change, but one that shows Zendesk is listening to the practical challenges of deploying AI in the wild.
The Reality Check: What’s Actually Changing?
Let’s cut through the jargon. For a while now, if you used Zendesk’s AI in mobile channels, every single message sent by the bot would have “Generated by AI” tacked onto the end. While transparent, it created a lot of noise and felt repetitive for the end-user. Even worse, for channels like SMS, that extra text was literally costing businesses money.
As of June 11th (with version 2.39 of the Mobile SDKs), Zendesk has rolled out a two-part solution:
- A New, Cleaner Label: The “Generated by AI” line item is gone. Instead, an “· AI” indicator is now integrated directly into the bot’s display name (for example, your bot “Zara” will now appear as “Zara · AI”). This makes it clear from the outset that the user is interacting with an AI, without repeating it in every single message.
- An Optional Disclaimer: The real game-changer is that you can now completely disable the per-message AI indicator. The new display name label is often sufficient for transparency, and this change gives you the control to remove the rest of the clutter.
This update applies automatically, but taking full advantage of it requires a small technical change, which we’ll cover below.
Our Expert Opinion: Why This Matters More Than You Think
On the surface, this seems like a minor cosmetic tweak. But from our perspective, this change has significant operational and financial implications. The announcement specifically calls out feedback from businesses using WhatsApp and SMS, and for good reason. The old system added up to 17 characters to every single AI-generated SMS message.
Let’s do some quick maths. A standard SMS segment is 160 characters. Adding 17 characters of repetitive legal text could easily push a message from one segment into two. If your bot sends three messages in an interaction, that could be the difference between paying for three SMS segments and paying for six. For a support organisation handling thousands of conversations a month via SMS, this seemingly small change could translate into hundreds or even thousands of euros in direct cost savings over a year.
Beyond the cost, this is also about Zendesk helping its customers navigate the complex landscape of AI regulation. The announcement mentions the EU AI Act, which requires transparency when users are interacting with AI. However, the law doesn’t necessarily mandate a disclosure on every single message. Zendesk’s new approach—a persistent label in the display name—is a much more elegant way to meet this compliance requirement. It respects the user’s intelligence and the flow of the conversation while still being upfront. It’s a mature solution to a modern problem.
How We’d Handle the Rollout: A Practical Action Plan
Here is what we’re advising our clients to do to take full advantage of this update. While the new display name format is automatic, turning off the per-message label requires a deliberate action.
| Phase | Action Item | Key Considerations |
|---|---|---|
| 1. Audit & Strategise | Review your current mobile bot conversations. | Does the per-message label feel intrusive? Are you using high-volume SMS or WhatsApp channels where character counts matter? Decide if you want to disable the per-message label. For most, the answer will be yes. |
| 2. Technical Implementation | Coordinate with your mobile development team to update your integration. | This isn’t a toggle in the Zendesk Admin Centre. Your developers need to set showAIDisclaimer to false using the Update Integration API for your iOS and/or Android integration. Ensure your team has planned for this update. |
| 3. Review Bot Persona | Check your bot’s new display name. | With “· AI” now part of the name, does it still look and feel right? A name like “Support Bot · AI” is clear. A more human-sounding name like “Chloe · AI” also works well. It’s a small but important branding touchpoint. |
| 4. Test & Validate | Once your SDK is updated to v2.39 and the API change is deployed, perform end-to-end testing. | Verify on both iOS and Android that the new display name appears correctly and that the old per-message label is gone. Ensure the conversation flow feels natural and less cluttered. |
Final Thoughts
This AI label redesign is a perfect example of a thoughtful, iterative improvement. It’s not a flashy new feature that will dominate a keynote, but it’s a practical change that solves real problems for customers. It reduces conversational friction for end-users, cuts operational costs for businesses, and provides a more sophisticated way to handle regulatory compliance. We’re pleased to see Zendesk refining the small details that make a big difference in the quality of day-to-day customer interactions.
