Five9: Bridging the CX Divide: Consumer Expectations for AI in Customer Service

The integration of artificial intelligence (AI) into customer experience (CX) operations is accelerating, with nearly all contact centers either piloting or implementing AI. However, a significant gap exists between CX leaders’ AI adoption strategies and consumer expectations. A recent study conducted by Five9 and Hanover Research in April 2026, surveying 3,000 consumers and 600 business decision-makers across the US, UK, and Germany, reveals critical insights into consumer trust, frustration points, and preferences regarding AI in customer service. For senior marketing and CX leaders, understanding these findings is essential to developing AI strategies that foster trust, reduce friction, and deliver measurable value.

The Imperative of Trust and Transparency in AI Adoption

Consumer trust in AI-powered CX is not universal. The research highlights a clear need for enterprises to build and maintain trust through transparency and the assurance of human support.

The Trust Gap and Transparency Mandate

Consumers exhibit a significant trust gap regarding AI’s ability to resolve issues independently. Over half of consumers (55%) trust AI to resolve their issue when a clear human option is available, compared to only 26% when a human option is absent. This differential, more than double, underscores the psychological safety a visible “human escape hatch” provides. Furthermore, transparency is paramount: 71% of consumers consider it important or very important to know if they are interacting with an AI bot or a person. For the Silent Generation, this preference rises to 91%.

This trust deficit extends to data security. Approximately 29% of consumers do not trust that information provided to AI agents is kept secure and private. While Millennials are the most trusting demographic in this regard (55%), older generations express greater skepticism. This indicates that perceived lack of trust is less about AI technology itself and more about the perceived gaps in an organization’s data governance and security policies.

What this means for leaders: Enterprises must integrate transparent disclosure mechanisms into their AI-driven CX. This includes clear identifiers for AI interactions and readily available options to escalate to a human agent.

• What to do:
• Implement clear AI disclosure: Proactively inform customers when they are interacting with an AI bot or virtual agent (e.g., “You are now speaking with our AI assistant”). This can be a visual cue in chat interfaces or an auditory statement in voice interactions.
• Guarantee human escalation paths: Ensure every AI interaction includes an explicit and easily accessible option to connect with a human agent. This should not require multiple prompts or navigation through complex menus. Monitor escalation rates and time-to-human transfer SLAs (e.g., target 30 seconds or less).
• Communicate data governance policies: Clearly articulate how customer data is handled, secured, and used in AI-powered interactions. Publish and promote these policies on corporate websites and within CX channels.
• Train AI with “humantic data”: Leverage successful human agent interactions to train AI models, focusing on empathetic responses, complex problem-solving, and efficient resolution. This improves AI capability and builds trust over time.
• What to avoid:
• Obscuring AI presence: Do not intentionally mislead customers about AI involvement, as this erodes trust and damages brand reputation.
• Hiding human escalation: Avoid making it difficult for customers to reach a human, which generates frustration and negatively impacts Customer Effort Score (CES).
• Vague security statements: Generic assurances about data security are insufficient; provide concrete details about encryption, data privacy protocols, and compliance (e.g., GDPR, CCPA).

Addressing AI Adoption Realities and Breaking Points

Despite the industry’s rapid adoption, consumers are not universally “AI-first.” Many use AI reluctantly, often reaching a “breaking point” that triggers escalation to a human agent.

The Reality of AI Usage and Escalation Triggers

Only 20% of consumers use AI for customer service “most of the time” or “always,” indicating that widespread AI exuberance among CX leaders does not fully align with consumer behavior. A significant segment of consumers (48%) actively tries to avoid using AI when possible, with this sentiment increasing with age. When consumers do engage with AI, 54% find the experience frustrating.

The critical “breaking point” for AI interactions is misunderstanding and the resulting handoff friction. A remarkable 92% of consumers give AI fewer than 10 minutes to resolve an issue before escalating to a human. The primary reason for this escalation, cited by 46% of consumers, is the AI bot repeatedly misunderstanding their query. This friction is compounded by poor handoffs: 35% of consumers report having to repeat themselves “often” or “always” when transitioning from an AI agent to a human. This leads to lower satisfaction, with AI and chatbot channels posting just 46% positive satisfaction, compared to 68% for phone interactions.

What this means for leaders: Focus on enabling seamless, context-rich handoffs from AI to human agents. Prioritize AI use cases where accuracy and rapid resolution can be confidently delivered, reserving complex or high-stakes interactions for human oversight or direct human engagement.

• Operating Model and Roles:

• AI orchestration layer: Implement a CX orchestration platform that maintains a unified customer view across all channels and interaction types. This system must capture all AI interaction context (transcripts, attempted resolutions, recognized intent, sentiment) and pass it seamlessly to human agents upon escalation.

• Agent Assist Tools: Equip human agents with AI-powered agent assist tools that surface relevant customer information and AI interaction history in real time. This reduces Average Handle Time (AHT) and improves First Contact Resolution (FCR) by eliminating the need for customers to repeat information.

• AI Trainer/Optimizer Role: Establish dedicated roles or cross-functional teams responsible for monitoring AI performance, analyzing escalation reasons, and continuously improving AI models based on “humantic data.” This involves red-teaming AI interactions and refining intent recognition and response generation.

• Defined Escalation Thresholds: Set clear performance thresholds (e.g., three consecutive misunderstandings, specific keyword triggers like “speak to a human”) for AI to automatically escalate to a human agent.

• What to do:

• Optimize AI-to-human handoffs: Invest in CRM and contact center software that preserves full interaction context, ensuring human agents have immediate access to the AI conversation history and customer data upon transfer.

• Prioritize simple, routine tasks for AI: Deploy AI for password resets and order/account status checks where AI alone can handle 38% and 35% of cases respectively. These tasks have a lower risk profile and higher consumer acceptance for AI-only resolution.

• Monitor FCR and CES for AI channels: Track these metrics closely. A low FCR or high CES for AI interactions indicates underlying issues with AI model accuracy or handoff processes. Aim for FCR rates above 70% for AI-handled tasks and a CES below 3.0.

• What to avoid:

• Over-automating complex issues: Avoid deploying AI for sensitive identity verification, charge disputes, legal compliance, or high-stakes escalations without robust human oversight (e.g., 60% of consumers prefer human-only for high-stakes escalation, Five9, 2026, p. 8).

• Ignoring AI performance metrics: Do not solely focus on containment rates. A high containment rate is detrimental if it leads to high frustration and customer churn due to unresolved issues.

• Fragmented data systems: Avoid deploying AI solutions that operate in silos, unable to access or contribute to a unified customer profile, which exacerbates handoff friction.

Tailoring AI Strategy for Diverse Audiences

Consumer sentiment toward AI is not monolithic; it varies significantly by generation and geography, necessitating a segmented approach to AI implementation.

Generational and Geographical Divides

Generational preferences play a crucial role in AI acceptance. Millennials are the most accepting demographic, with 55% finding AI useful for customer service. In contrast, the Silent Generation is the most human-reliant, with 55% believing AI offers no customer service benefit. They predominantly prefer phone interactions (two-thirds) and nearly three-quarters (73%) explicitly switch away from AI for complex issues. Gen Z, despite being the youngest, shows mixed sentiments; 1 in 10 say AI has no benefit, yet they are more open to its use as models evolve.

Geographical differences are also pronounced. German consumers are notably more bullish on AI, with only 9% believing AI has no benefits, compared to 23% in the UK and 18% in the US. Germans are also more likely to use AI options (23% “most of the time” or “always”) than US (18%) or UK (19%) consumers. Conversely, the UK expresses the highest concern for AI transparency (76% deem it important) and experiences the most frustration with AI (60% find it frustrating), along with the highest rate of repeating information during AI-to-human handoffs (42%).

What this means for leaders: A “one-size-fits-all” AI strategy is ineffective. CX leaders must develop nuanced AI playbooks that consider demographic and regional specifics, leveraging AI to personalize interactions and channel choices.

• Governance and Risk Controls:

• Audience segmentation policy: Establish clear policies for segmenting AI use cases and communication strategies based on generational, geographical, and potentially industry-specific preferences.

• Dynamic Channel Orchestration: Implement systems that can dynamically route customers to preferred channels (e.g., phone for Silent Generation, email for German customers, Five9, 2026, p. 11) or AI experiences based on their profile and interaction history.

• Consent Management: Ensure clear consent mechanisms are in place for data usage by AI, especially for sensitive interactions. This includes explicit opt-in/opt-out options for AI-driven personalization.

• Regional AI Compliance: Adhere to local regulations regarding AI transparency and data privacy (e.g., German data protection laws are stringent).

• Immediate Priorities (First 90 days):

• Audit current AI deployments: Evaluate existing AI interactions for transparency, human escalation paths, and handoff quality. Identify high-friction areas.

• Map AI capabilities to customer segments: Overlay current AI capabilities with generational and regional consumer preferences. Pinpoint areas where AI is overused or underutilized for specific segments.

• Develop a “human-in-the-loop” framework: Define where human oversight is essential for AI interactions (AI + human oversight) and where only human intervention is appropriate (Human only), especially for tasks like sensitive identity verification or legal compliance.

• Pilot improved handoff protocols: Implement and test enhanced handoff procedures that ensure full context transfer from AI to human agents in a high-priority segment (e.g., for Baby Boomers or in the UK market, where handoff issues are most prevalent).

• What ‘good’ looks like:

• High CSAT/NPS for AI-assisted interactions: Scores for AI channels approach or match human-agent interactions, indicating effective AI support.

• Low complaint rates related to AI: Minimal customer complaints about AI misunderstanding, frustration, or lack of human access.

• Demonstrable ROI for AI: AI deployments show measurable improvements in efficiency (e.g., reduced AHT for simple tasks, agent productivity gains) and customer outcomes (e.g., higher FCR, increased renewals/conversions).

• Segment-specific success metrics: Achieve tailored performance goals for different customer demographics and regions, reflecting their unique preferences and trust levels.

Summary

The widespread adoption of AI in CX presents a transformative opportunity, but its success hinges on a nuanced understanding of consumer perceptions and expectations. CX leaders must prioritize building trust through unwavering transparency, ensuring clear human escape hatches, and securing customer data with robust governance. Furthermore, addressing consumer “breaking points” by optimizing AI-to-human handoffs and reserving AI for appropriate, low-risk tasks will reduce frustration and improve satisfaction. Finally, a segmented strategy, accounting for diverse generational and geographical preferences, is critical for tailoring AI experiences that truly resonate with customers. By focusing on these principles, enterprises can move beyond mere AI implementation to achieve sustained, measurable improvements in customer experience and loyalty.

Source: Five9. (2026). What Consumers Really Think About AI in Customer Service. [eBook]