From experimental to essential: The state of AI in CX 2025 report

AI in CX in 2025: Key takeaways

Methodology: For this study, CCW Europe surveyed 100+ thought leaders from the CCW Europe community—business pioneers across customer support services, operations, product management, and more—to assess their approach to adopting AI for CX.

Here are the key takeaways:

• Top drivers of interest: Customer satisfaction (71%), workforce productivity (63%), and cost reduction (56%).

• Biggest barriers to adoption: Data security (44%), lack of skills (44%), and cost pressures (39%).

• Top AI technologies: Conversational AI (63%) such as chatbots, knowledge management (40%), and workflow automation (38%).

• Measures to build customer trust: 44% are testing AI features in smaller groups pre-rollout, 38% offer a human alternative, 21% explain how AI decisions are made.

• 95% plan to increase AI spend over the next 12 months, despite mixed outcomes (14% highly successful, 29% mixed results).

• The majority (40%) of organizations consider AI a core strategic pillar, 29% say it has secondary support, and 31% are in the experimental phase.

• Only 22% have live deployments, while 31% remain stuck in pilots.

1. Investment driven by experiential and operational goals

The study shows that three priorities dominate the investment behind AI in CX:

• Improving customer satisfaction (71%)

• Boosting workforce productivity (63%)

• Reducing costs (56%)

The data reflects that AI represents an experience enhancer and an efficiency engine for enterprises. These investments acknowledge the growing expectation among customers for faster, more cohesive interactions across both online and offline channels. This aligns with a wider shift to customer centricity as a competitive differentiator, helping organizations to meet these rising demands at scale while also increasing operational efficiency and agility.

2. Responsible AI presents varied challenges

The study finds that various roadblocks and challenges are top of mind for enterprises as they try to implement AI safely and responsibly. These include:

• Data security risks (44%)

• Lack of internal skills/knowledge (44%)

• Cost pressures (39%)

• Hallucinations/inaccurate outputs (36%)

• Integration barriers and data silos (31%)

Transformation implies a certain amount of turbulence, and future-facing organizations are buffeted by varied challenges they must address to unlock lasting value from AI.

On the technical side, organizations report facing a double hurdle as AI functionalities come online: being vulnerable to data breaches, and also ill-equipped to handle the breadth of sensitive data needed to fuel AI systems. On the human side, a skills gap can leave organizations rudderless and running behind, forced to grapple with complex technical challenges while trying to avoid budget overruns and demonstrate ROI for AI.

3. Spending is climbing despite mixed outcomes

Despite the headwinds, almost every organization (95%) plans to increase its investment in CX-focused AI over the next 12 months. Notably, none said they would reduce spend.

The predominant share of organizations (59%) expect to modestly increase AI expenditure in the next year, with a practical focus on scaling practical use cases, optimizing current applications, and strengthening AI governance protocols. However, a full (36%) anticipate a significant uplift, shifting from experimentation to expansion out of pilots into broader integration.

Current results with AI in CX are uneven:

• 14% highly successful

• 23% moderately successful

• 29% mixed results

• 34% too early to tell

As in other industries, momentum is picking up around AI in CX. In 2024, only 38% of surveyed C-suite CX executives rated AI as a priority. In 2025, that figure climbed to 54%.

4. Budgets lean into conversational AI

With investment set to increase, organizations say they plan to incorporate various AI technologies. These follow a few trendlines:

• Conversational AI remains the cornerstone. While 63% of respondents say they have already invested in conversational AI (chatbots, voice AI agents)—a full 56% plan to expand or double down. Leaders are evolving these tools for more natural, context-aware engagement, underscoring their pivotal role in managing and triaging customer touchpoints at scale.

• AI quality/insight tools are gaining. Nearly a third of CX leaders (36%) say sentiment analysis, language assist (33%), and conversation analysis (31%) are near-term priorities. Still relatively rare today, these technologies are central to building more emotionally intelligent systems, essential for human-centric AI in CX.

• More advanced capabilities lag behind. Fewer companies are prioritizing predictive analytics (21%), intelligent call routing (18%), and workforce management/forecasting (21%) for immediate investment. These require stronger data maturity and integration, and are often activated once foundational AI maturity is in place.

5. The road to customer confidence is uneven

As AI becomes more embedded in customer-facing processes, user trust is emerging as the currency of successful adoption and sustainable value. The survey shows organizations are still in the early stages of fostering trust in AI with customers, with more than a quarter (26%) saying they aren’t taking any specific steps to build customer confidence in their AI use.

Among those that are focused on building customer trust, strategies vary in scope:

• Cautious rollouts are prevalent. The most popular approach (44%) is to test AI features with small groups before scaling. This lowers immediate risk but rarely addresses deeper concerns about AI transparency or governance.
  

• Human fallbacks are common. Roughly 38% provide customers the choice to engage with a human instead of AI alone—an interim step that helps close trust gaps.

• Transparency efforts are limited. Only 21% explain to users how AI decisions are made, 18% disclose when AI is at play, and just 13% clarify data usage or processing practices upfront.

• Customer engagement is underutilized. A minority (13%) collect and incorporate customer feedback on AI interactions, and just 10% educate their customers.

As AI becomes table stakes, organizations that lean into transparency and active customer involvement as a means to competitive CX will be best positioned to earn loyalty in an AI-driven future.

6. Employee uptake: The overlooked factor

Employee adoption remains one of the most under-prioritized aspects of AI in CX. The study shows:

• Only 8% of organizations fully prioritize employee adoption.

• 15% have largely prioritized employee uptake.

• 40% say they’ve addressed adoption minimally—or not at all.

For all the focus on technology, many enterprises overlook the human side of AI transformation. The study confirms recent research from Writer that finds some Gen Z and Millennial employees are actively resisting or even sabotaging their company's AI initiatives. Without clear communication, training, and empowerment, employees are more likely to resist adoption than embrace it.

As Kevin Chung from Writer noted on Sendbird’s MindMakers podcast, adoption is often the “last mile” where AI projects break down. Making employee engagement and change management a top priority is often the difference between stalled pilots and scalable success.

7. Reactive compliance practices prevail

The regulatory landscape around AI is evolving quickly. Organizations must navigate a patchwork of state, regional, and industry-specific mandates for how AI systems should be designed, deployed, and monitored to ensure their safe and responsible operations.

The report shows that most organizations are taking a reactive approach to compliance:

• 54% rely on internal legal or compliance teams to track mandates.

• 33% participate in industry events or working groups.

• Only 26% receive regular legal briefings or vendor updates.

• Just 18% engage external consultants, and only 10% use regulatory sandboxes—missing opportunities to shape and test governance in practice.
  

This reactive posture carries risks. For instance, data privacy concerns extend beyond model training and outputs to the enterprise data fueling AI—especially when it contains personally identifiable information (PII) from users.

In this environment, a structured approach to AI oversight, data protection, and transparency is not just a compliance requirement—it’s a strategic enabler of trust. Organizations that embed these principles from the outset will be best positioned to scale AI safely and with confidence.

8. The gap between ambition and execution

While 88% of the surveyed CX organizations report some form of AI deployment—which exceeds the estimates of 78% deployment across industries—many deployments remain at a limited scale:

• 35% have reached moderate maturity.

• 33% are still operating at a small scale.

• 20% are at the pilot stage.

• Only 12% report full enterprise-wide deployment.

This uneven maturity maps directly onto the varied adoption models in play. Of those surveyed, the most common approach (40%) was a decentralized model where individual teams manage AI independently. Less than a quarter (23%) of organizations have a fully centralized function for AI adoption and management, while the rest (37%) fall somewhere in between, with partially centralized ownership. 

As it becomes clear that more than technology is needed to unlock value from AI, organizations are seeing that enthusiasm has less of an impact on AI maturity than how they structure their AI ownership, governance, and trust.

Those who build on a strong foundation of centralized oversight for AI—while empowering each business unit to innovate in accordance with best practices on AI observability, control, and responsibility—are most likely to convert pilots into scaled solutions that generate sustainable value.

AI in CX: From experimental to essential

​​AI is advancing at a pace that rivals any past technology shift in business, and CX has perhaps the most to gain of all. With that speed, enterprise CX leaders face a choice: play catch-up or lead the way.

While the playbook is still being written, one thing is clear: success requires investing in enterprise-ready capabilities, making customer confidence a priority, and embedding trust and governance from the start.

Rather than “bolting on” AI, CX leaders who treat AI as a powerful strategic lever that necessitates an integrated approach are poised to unlock significant gains both for customers and operations teams.

To access all the insights on the state of AI in CX in 2025, download the full report. (Coming soon!)