Enterprise adoption of artificial intelligence in customer experience is advancing faster than organizations can connect AI systems to the people, data and workflows required to resolve customer requests, according to new research commissioned by Talkdesk.

The report, “The State of Agentic Automation in CX,” found that 98% of surveyed organizations have deployed AI somewhere in the customer journey, but only 15% combine agentic AI with cross-departmental orchestration. Just 5% said they can quantify AI's impact on business outcomes.

The findings point to a widening gap between AI deployment and operational execution. For customer experience leaders, the challenge is increasingly less about acquiring AI capabilities and more about integrating them into business processes.

NewtonX conducted the survey for Talkdesk among 252 director-level and above decision-makers and influencers responsible for customer experience, IT, operations or AI strategy. Respondents represented mid-market and enterprise organizations across North America, EMEA, Latin America and Asia-Pacific, with industries including healthcare and financial services.

AI Adoption Is Outpacing Operational Readiness

The report suggests that enterprises have moved rapidly from AI experimentation to deployment, but many implementations remain isolated.

Specialized AI agents are used by 64% of respondents, yet only 35% retain customer context when information moves between systems. Without that continuity, an AI system or human employee may have to reconstruct the customer's history before completing an interaction.

Technical infrastructure is another constraint. Forty-five percent of organizations identified disconnected systems as a roadblock, while 44% cited legacy infrastructure. Those limitations leave nearly 80% of organizations operating with 10 or fewer AI automations, according to the research.

The consequences extend to human employees. Talkdesk reported that human customer-service agents spend an average of 28% of their time switching systems, re-entering information and searching for customer context when AI-enabled workflows cannot complete a task.

This creates a potential paradox: enterprises can invest in more AI tools while retaining the manual processes those tools were intended to reduce.

The Orchestration Gap

The report defines orchestration as the ability to coordinate AI agents, employees, data and workflows across enterprise systems.

That distinction matters because customer problems rarely remain inside one application or department. A billing question may require information from a CRM system, payment platform and customer-service application. A retention request may involve customer history, product usage and a commercial decision.

An AI model can generate an answer without necessarily completing the underlying business process.

Talkdesk's findings indicate that 85% of organizations lack the orchestration capabilities required to connect these components and deliver end-to-end resolution.

Trust is another barrier. Fifty-two percent of respondents identified confidence in AI decisions as a primary concern. Meanwhile, 94% of organizations operate without AI-assisted knowledge management, potentially limiting the information available to agents when they make or recommend decisions.

The result is a technology stack in which AI may be capable of handling individual tasks but is not yet consistently positioned to own complete workflows.

AI Agents Become Part of the Workforce

The research also reflects a conceptual shift in enterprise AI adoption. Nearly one in five organizations already view AI agents more as a form of labor than as conventional software.

At the same time, 99% of respondents said a hybrid workforce combining AI and people provides value.

That combination creates new operational questions for enterprises. Traditional software is generally managed through application ownership, IT administration and user permissions. AI agents can perform tasks, make recommendations and potentially act across multiple systems, creating a need for new governance, monitoring and accountability models.

The research found that organizations with the highest customer-experience automation maturity are more than 10 times as likely to operate AI and human employees as a unified workforce.

The implication for marketing and customer-experience organizations is significant. AI agents are increasingly becoming participants in customer journeys rather than merely tools used by employees.

Market Landscape: From AI Tools to AI Operations

The customer-experience software market is moving toward a more integrated model in which conversational AI, automation, analytics, CRM data and workflow management increasingly overlap.

Major technology vendors including Salesforce, Microsoft, Google and ServiceNow are investing in AI agents and enterprise workflow automation.

The competitive issue is therefore shifting. Access to a capable foundation model or individual AI assistant is becoming less differentiated as enterprises evaluate how effectively those technologies work with proprietary data, business rules and existing applications.

Talkdesk's research reinforces that point. Organizations combining agentic AI with cross-departmental orchestration were four times more likely to report major improvements in customer satisfaction or NPS.

Higher-maturity organizations were also nearly twice as likely to automate revenue-oriented use cases such as churn prediction and personalized recommendations.

Strategic Outlook

The next phase of enterprise CX automation is likely to focus on measurable workflow outcomes rather than AI deployment counts.

One metric to watch is autonomous resolution. The report found that 38% of organizations at the highest maturity level autonomously resolve more than 40% of customer issues, while none of the organizations in the lowest maturity tier reached that threshold.

Demand is expected to increase. Eighty-three percent of surveyed organizations expect autonomous issue-resolution rates to rise over the next two years.

For enterprises, that trajectory raises three priorities: connecting fragmented systems, establishing governance for AI agents and creating reliable measurement frameworks.

The last of these may be particularly important. If only 5% of organizations can currently quantify AI's business impact, increasing deployment without better measurement could make technology spending harder to evaluate.

The central competitive advantage may therefore move from simply deploying AI to building an operating environment in which AI can safely complete work across systems and departments.

Top Insights

• 98% of surveyed organizations have deployed AI somewhere in the customer journey.
• Only 15% combine agentic AI with cross-departmental orchestration.
• Just 5% can quantify AI's business impact.
• 35% retain customer context across systems.
• Human agents lose an average of 28% of their time to system switching, data re-entry and context searches.
• 52% cite trust in AI decisions as a primary concern.
• Organizations combining agentic AI and orchestration are four times more likely to report major CSAT or NPS gains.
• 83% expect autonomous issue resolution to increase within two years.

Get in touch with our MarTech Experts