Most CX Leaders Are Still Buying for the Last AI Cycle

Salesforce, HubSpot, Genesys, and Microsoft are all pointing to the same shift, where orchestration, governance, and execution depth matter more than front-end polish.

4
CX AI Control Room
AI & Automation in CXFeature

Published: July 7, 2026

Rob Wilkinson

A lot of the CX market is still evaluating AI at the surface layer. The conversation is often dominated by interface quality, conversational fluency, and whether the assistant feels smart enough to impress in a demo. That lens is becoming less useful by the week. The more important shift now is happening underneath the experience layer, where orchestration, workflow access, governance, and execution depth are starting to define which platforms can actually operationalize AI.

That is the real story running through this week’s moves from Salesforce, HubSpot, Genesys, and Microsoft. On paper, the stories look different. One is about commerce, one about CRM, one about contact center orchestration, and one about secure infrastructure. But taken together, they suggest the market is moving into a phase where AI has to do more than present intelligence. It has to prove it can act inside enterprise systems without creating more operational complexity than value.

The Visible AI Layer Is Becoming Less Important Than the Operating Layer

For the last two years, AI in CX has largely been discussed through the lens of assistance. Vendors improved agent guidance, automated summaries, reduced manual effort, and made self-service experiences more natural. Those changes mattered, but they did not fundamentally answer the harder enterprise question: what happens when AI has to move from helping work happen to actually doing parts of the work itself?

This week’s stories suggest that question is now moving to the center of the market. The most meaningful progress is no longer just about how AI sounds. It is about whether it can connect to live systems, work across fragmented processes, and stay inside commercial and governance boundaries.

That shift is especially visible in commerce. Salesforce’s Agentforce Commerce launch is not just another AI assistant story. It is an attempt to make AI transactionally useful in a live buying environment. The bigger strategic message is that AI may shape discovery, but brands still need conversion to happen where they retain control. As Nitin Mangtani, EVP & GM of Agentforce Commerce at Salesforce put it:

“The brands that win will have their Shopper Agent live on their own properties for the 2026 shopping season.”

That line matters because it reveals where Salesforce thinks the next battle will be won. Discovery may increasingly happen through third-party AI surfaces, but commercial control still depends on what the brand can operationalize in its own environment.

Orchestration Is Emerging as the Real Competitive Fault Line

If Salesforce’s story is about transaction-level usefulness, Genesys’ acquisition of Pinkfish highlights the deeper problem vendors are trying to solve. Most enterprise AI systems still struggle at the point where a customer interaction needs to turn into coordinated action across multiple systems. Understanding intent is one thing. Resolving the issue across CRM, ERP, billing, and order systems is something else entirely.

That is why orchestration now matters more than the front-end bot itself. The market is beginning to recognize that the real competitive advantage may sit in the layer that connects AI to business logic, permissions, and workflow execution. Glenn Nethercutt, EVP and CTO of Genesys framed the shift clearly:

“Agentic AI is moving customer experience from assisted engagement to governed execution.”

That is a concise way of describing the market transition now underway. The next AI battleground is whether platforms can coordinate meaningful action under enterprise-grade guardrails.

Passive Platforms Are Starting To Look Exposed

HubSpot’s acquisition of Warmly points to a related pressure inside CRM. Systems of record are increasingly vulnerable if they cannot become systems of signal and action. In a market where businesses are under pressure to identify buying intent earlier, personalize outreach faster, and reduce manual effort without adding headcount, passivity starts to look like a strategic weakness.

That is why the Warmly move matters. It signals that CRM vendors now need to prove they can surface signals, trigger action, and support revenue execution more directly. They cannot just document activity after the fact.

The strategic urgency behind that shift came through when Angela DeFranco, GM, VP of Product at HubSpot argued:

“The gap between building demand and winning deals is one of the hardest problems in GTM.”

The bigger point is what the acquisition reveals. HubSpot appears to believe the gap between signal and action is now strategically important enough that waiting to close it through slower internal development would carry its own risk.

Governance Is Becoming Product Infrastructure

Microsoft’s managed MCP server for Dynamics 365 Commerce may be the least theatrical story of the group, but it may also be one of the most revealing. The announcement signals that governance is no longer something the market can leave at the edge of the AI narrative. If enterprises want AI agents to interact with live commerce systems, they need controlled access, explicit authorization, visibility, and a standardized way to govern those actions.

That shift matters because it makes governance less abstract. It is becoming architecture.

In an assessment of the broader challenge, Heath Ramsey, GVP of Product Management, AI Platform at ServiceNow warned:

“The gap is the governance and the visibility, especially within the CX space.”

That is a sharper way to frame where the market now stands. The limiting factor is increasingly whether enterprises can trust the mechanisms by which it does it.

What Enterprise Buyers Should Infer Now

The most important lesson this week is that the market is becoming harder to read from the surface. If buyers keep focusing on what the AI layer looks like, they risk missing where the real competitive advantage is forming.

The better questions now are architectural. Where does orchestration live? How does the system move from intent to action? What permissions control that action? What happens when a workflow fails? How much of the vendor’s story is live, governed capability, and how much is narrative momentum racing ahead of operational reality?

That is the shift serious CX leaders need to pay attention to now. The risk is buying the wrong layer of the stack while the real market advantage forms somewhere less visible, but much more defensible.


Join the conversation: Join our LinkedIn community (40,000+ members): https://www.linkedin.com/groups/1951190/
Get the weekly rundown: Subscribe to our newsletter: http://cxtoday.com/sign-up

Agentic AI in Customer Service​Agentic AI SoftwareAI AgentsAutonomous Agents
Featured

Share This Post