IT departments now control contact center AI procurement decisions, with fewer than one in five buying choices made by the contact center itself. This shift reflects the technical complexity embedded in modern AI tooling—conversational AI platforms, agent assist systems, and orchestration engines require infrastructure expertise, security validation, and integration oversight that traditionally sat outside CX teams' remit. The consequence is a fundamental misalignment: IT prioritises system resilience, data governance, and vendor consolidation, whilst contact center leaders care about agent productivity, customer resolution rates, and operational flexibility. Zendesk's outcome-based pricing model and Five9's Acqueon acquisition both signal vendor confidence in AI-driven automation, yet neither addresses the procurement bottleneck. When IT holds veto power over tools that directly affect frontline performance, how do CX teams ensure their operational requirements actually shape the technology stack rather than merely rubber-stamping IT's infrastructure decisions?
The trust deficit compounds this structural problem. Despite years of investment—Zoom acquiring Solvvy, Zendesk embedding conversational AI into its CRM, Cisco acquiring Robust Intelligence for AI security—customer sentiment toward AI-driven service remains negative. The sources point to poorly tested conversational journeys, missed micro-intents, and gaps between what AI can technically do and what it reliably does in production. Intercom's rebrand to Fin acknowledged that AI agents had become the core business, yet the company's CEO admitted that newer competitors succeeded partly because they carried no legacy baggage—they didn't need to convince anyone of a new market position. This observation cuts both ways: established vendors like Zendesk and Cisco have the resources to build robust AI, but they're also burdened by existing customer bases running older systems. Smaller vendors entering the space face no such constraint, which raises a harder question for support leaders already running Agentforce or Zendesk agents: if IT procurement now controls AI buying, and trust in AI remains fragile, what recourse do you have when the chosen platform underperforms against customer expectations?
The infrastructure layer adds another layer of risk. Vonage's 36-hour SMS outage exposed how single points of failure in cloud infrastructure can disable critical customer workflows, and the IMF's warning about Claude and AI-driven cyber threats suggests that as AI becomes embedded deeper into financial CRM systems, the attack surface expands faster than defences can scale. Zendesk's shift to outcome-based pricing—charging only for AI-resolved issues—transfers risk from vendor to customer, but only if the AI actually resolves issues reliably. For teams managing contact centers, this means the procurement decision made by IT today will determine not just your technology stack, but your operational resilience, your ability to respond to customer trust issues, and ultimately your exposure to both AI failures and security incidents. The real question isn't whether IT should buy contact center AI; it's whether IT procurement processes are equipped to evaluate AI quality assurance, testing rigour, and customer trust metrics alongside infrastructure requirements.
BIG CX NEWS from Zoom, Zendesk, and Avaya cxtoday.com
Big CX News from Vonage, Claude, Fin (Intercom) & Cisco cxtoday.com
Big CX News from Zendesk, Cisco, Oracle & Five9 CX Today
BIG CX NEWS from Avaya, Zendesk, and Qualtrics CX Today
IT Now Calls the Contact Center AI Buying Shots: So What? cxtoday.com