Safely manage your Zendesk from the AI assistant you already use, via the Deltastring MCP. Beacon configuration platform
← Back to news

Announcing AI Label Redesign on Mobile SDKs

Zendesk

Zendesk has redesigned how AI disclosure appears in Mobile SDKs, moving from a repetitive per-message "Generated by AI" label to an integrated indicator within the bot's display name (e.g., "Zara · AI"). The change addresses a practical friction point: the previous labeling approach added up to 17 characters per message on SMS channels, directly inflating costs, whilst creating visual clutter particularly on bandwidth-constrained platforms like WhatsApp and SMS. Critically, Zendesk has made this optional—teams can now disable per-message labeling entirely via the Update Integration API—whilst maintaining compliance with transparency regulations including the EU AI Act. The update rolls out automatically in Mobile SDK version 2.39 on 11 June with no mandatory action required.

The implications cut across cost efficiency and regulatory risk management. For teams operating high-volume SMS or WhatsApp channels, the 17-character reduction per message translates to material savings at scale; a support team sending 10,000 AI-generated SMS responses monthly would recover roughly 170,000 characters of transmission cost. Yet the optional nature of per-message labeling introduces a compliance consideration: whilst the EU AI Act does not mandate per-message disclosure, enforcement bodies may interpret "clear understanding" differently across jurisdictions. Teams should audit their current disclosure strategy against their specific regulatory landscape before disabling the label entirely. The redesign also signals how Zendesk is calibrating its AI transparency approach—moving away from blanket labeling toward contextual, channel-aware disclosure—which raises a broader question for CX leaders: as AI disclosure regulations mature, will the industry converge on similar pragmatic solutions, or will fragmented compliance requirements force teams to maintain multiple disclosure strategies across regions?