Twilio's 2026 research exposes a critical implementation gap in Asia Pacific customer service automation: whilst 84% of businesses believe their AI agents effectively recognise returning customers and maintain conversational history, 70% of consumers report having to restart conversations from scratch. This 62-point perception chasm reflects a systemic failure in data architecture rather than AI capability alone. The problem compounds at handoff points—65% of consumers must repeat information when transferred from bot to human agent—because customer histories remain fragmented across disconnected systems. For CX teams already invested in platforms like Zendesk or Salesforce, this signals that AI deployment without unified customer data infrastructure creates friction that directly drives brand abandonment, with 5% of respondents leaving brands entirely after poor AI interactions.
The commercial stakes intensify as Asia Pacific businesses plan to escalate AI handling of customer service from 52% of interactions today to 65% by 2027, yet consumers simultaneously demand greater transparency and control. Seventy percent expect AI agents to identify themselves at conversation start—a disclosure rate only 22% of businesses currently meet—whilst 57% want human approval for AI actions and 51% want visibility into what data systems can access. This creates a paradox: organisations are accelerating automation deployment whilst their customers are signalling that trust erodes when AI operates without clear boundaries or human oversight. For support leaders managing these transitions, the question becomes whether expanding AI capacity without first resolving data silos and transparency gaps will simply scale frustration rather than efficiency.
The research suggests that consumer comfort with AI for routine tasks (86% willing to use it for appointments, 85% for returns) coexists with deep scepticism about AI handling complex, history-dependent interactions. This distinction matters operationally: teams should audit whether their current data integration strategy—across legacy systems, third-party integrations, and knowledge bases—can actually support the contextual continuity consumers expect before increasing AI's scope. The 91% of organisations already building proactive AI workflows indicates the industry is moving faster than its infrastructure can support, creating a window where CX leaders who prioritise data unification now will gain competitive advantage over those who treat it as a post-deployment problem.
Poor AI memory is pushing Asia Pacific consumers to abandon customer service chats Tech Edition
Poor AI context is pushing Asia Pacific consumers to abandon customer service chats Tech Edition