The integration of Artificial Intelligence (AI) into customer-facing operations promises enhanced efficiency and personalized interactions. However, a significant gap remains between AI’s assistive capabilities and consumer readiness for AI to manage high-stakes transactions autonomously. A recent study by Expedia Group, The AI Trust Gap, highlights that while travelers embrace AI for discovery and planning, their preference for booking and critical support remains firmly with trusted brands, not AI platforms. This distinction underscores a critical challenge for senior marketing and CX leaders: building confidence in AI means prioritizing accountability, data governance, and robust operational models.
The AI Trust Gap: Distinction Between Discovery and Decision
The study, which surveyed travelers across the US, UK, and India, reveals a clear bifurcation in consumer AI adoption. Travelers are comfortable with AI for information gathering and preliminary planning but demonstrate significant hesitation when it comes to transactional activities.
- Preference for Trusted Brands: Sixty-eight percent of travelers prefer to book with a trusted travel company rather than directly through AI platforms. A substantial sixty-six percent explicitly state they do not trust AI to book on their behalf. This indicates that while AI can inform, it has not yet earned the authority to act independently in high-value transactions.
- AI as an Assistive Tool: Consumers are receptive to AI in a supporting capacity. The study found that fifty-three percent are comfortable with AI suggesting options, forty-two percent would use AI to monitor prices and determine optimal booking times, and forty percent would leverage AI for itinerary building. This highlights AI’s strength in information processing and recommendation generation.
- High-Stakes Transactions Require Trust: The travel sector exemplifies industries where customer errors or system failures carry significant financial and emotional costs. A single booking mistake or an issue during a trip can result in substantial monetary loss or a ruined experience. In such scenarios, travelers seek assurance that a reputable entity stands behind the transaction, capable of resolving unforeseen problems. AI chatbots, while adept at summarization, lack the operational infrastructure and established supplier relationships to address complex mid-trip changes or booking errors effectively.
Summary: The data confirms that AI’s role is currently perceived as a powerful enhancer for exploration and initial recommendations. However, for critical stages involving financial commitments and potential issue resolution, human-backed, trusted brands remain paramount. This is not a limitation of AI’s technical capabilities but rather a reflection of the inherent trust dynamics in consumer behavior.
Cultivating Trust Through Accountability and Control
The primary barriers to consumers entrusting AI with transactional authority stem from concerns about control, data privacy, and accountability. Enterprises deploying AI must address these foundational trust issues directly through transparent policies, robust data governance, and clear operational guardrails.
- Core Trust Concerns: The study identifies specific anxieties:
- Loss of control: Fifty-seven percent of travelers expressed concern over losing control when AI acts on their behalf.
- Data privacy: Fifty-seven percent cited data privacy as a major concern.
- Misuse of personal data: Fifty-six percent were worried about the misuse of their personal data. These concerns extend beyond the travel sector, resonating across financial services, healthcare, and retail where sensitive personal information and high-value decisions are involved.
- Building Brand Accountability: Trust is built on a brand’s established record of expertise, deep customer relationships, and end-to-end support. In an AI-augmented environment, this means clearly defining the scope of AI operations and maintaining human oversight for critical functions. For example, a financial services institution might use AI for initial credit score assessments but require human approval for loan disbursements, maintaining stringent thresholds for automated actions (e.g., “automated approval for credit limits up to $500, manual review for anything above”).
- Data Readiness and Governance: To mitigate data privacy concerns, enterprises must ensure robust data governance frameworks. This includes transparent consent mechanisms for data collection and usage, strict access controls, and compliance with regulations such as GDPR or CCPA. High-quality, structured data is crucial for AI accuracy and preventing misinformation, which can erode trust rapidly. For instance, in retail, maintaining accurate inventory data across all AI and human-assisted channels prevents customer frustration from misinformed recommendations or unavailable products.
What to do:
- Implement clear AI usage policies: Define when AI operates autonomously and when human intervention is required, especially for critical transactions (e.g., “AI provides initial quotes, human agent confirms booking”).
- Prioritize data privacy and security: Invest in encryption, anonymization, and regular security audits. Communicate data handling practices clearly to customers.
- Establish accountability frameworks: Designate clear escalation paths for AI-driven errors, ensuring customers know who to contact and what recourse they have.
- Develop high-quality, structured content: Ensure all data fed to AI models (product descriptions, pricing, policies) is accurate and up-to-date, preventing factual errors that undermine trust.
What to avoid:
- Over-automating high-stakes decisions: Do not allow AI to complete critical transactions (e.g., booking, fund transfers, medical advice) without explicit customer consent and an accessible human override.
- Vague data privacy statements: Avoid generic disclaimers. Be specific about what data is collected, why, and how it is protected.
- Ignoring AI-generated errors: Establish immediate error detection and resolution protocols. A single unaddressed AI error can significantly damage brand reputation.
Operationalizing Trust in AI: Roles, Metrics, and Integration
Integrating AI effectively while preserving customer trust requires a strategic approach to operational models, system integration, and performance measurement. CX leaders must design workflows that leverage AI’s strengths without ceding human accountability.
- Operating Model and Roles: Define specific roles for human agents in conjunction with AI. This is not about replacing agents but augmenting their capabilities. For example, in a telecom customer service center, AI can handle initial query routing and provide agents with comprehensive customer histories from CRM systems, reducing time-to-resolution (target: 20% reduction in average handle time). Agents then focus on complex problem-solving and empathetic interactions. Implement a “human-in-the-loop” model where AI outputs for sensitive queries are flagged for review.
- Seamless Integration: AI systems must be tightly integrated with existing enterprise platforms. This includes CRM, billing systems, ticketing platforms, and inventory management. Such integration ensures data consistency and allows AI to draw from a complete, accurate customer profile. For a B2B SaaS company, AI-powered chatbots can qualify leads (e.g., company size, industry, specific needs) using data from a Salesforce CRM, then seamlessly hand off to a sales development representative with pre-populated context.
- Governance and Risk Controls: Establish clear guardrails and thresholds for AI operations. This involves continuous monitoring, red-teaming exercises to identify potential biases or failure points, and a robust change management process for AI model updates. Define SLAs for AI system performance and clear escalation paths for incidents where AI behavior deviates from expected parameters (e.g., “if AI response sentiment falls below 60% positive, flag for human review and root cause analysis within 2 hours”).
- Measurement Beyond Containment: While AI can improve efficiency metrics like call containment or response speed, the ultimate measure of success is its impact on customer trust and business outcomes. Focus on metrics such as:
- Customer Satisfaction (CSAT)/Net Promoter Score (NPS): Track changes in these scores for AI-assisted interactions versus purely human interactions.
- First Contact Resolution (FCR): Ensure AI-powered tools contribute to higher FCR, indicating effective problem-solving.
- Customer Effort Score (CES): Evaluate if AI makes interactions easier or more complex for the customer.
- Complaint Rate: Monitor any increase in customer complaints related to AI interactions.
- Conversion and Renewal Rates: For sales or retention-focused AI, measure direct impact on business goals.
- Error Rate/Misinformation Rate: Track specific instances where AI provides incorrect or unhelpful information.
What ‘good’ looks like: A successful AI implementation demonstrates improved efficiency without compromising customer trust. A financial services firm might achieve a 15% increase in self-service inquiries handled by AI, alongside a sustained CSAT score of 85% for these interactions, and a <0.5% complaint rate linked to AI errors. Human agents are empowered with AI tools, leading to a 10% reduction in agent burnout and a 5% increase in complex case resolution efficiency.
The “AI Trust Gap” identified by Expedia Group serves as a crucial reminder for all enterprise leaders: AI’s true value is realized not merely through its technical prowess, but through its integration into operations that are fundamentally built on trust, transparency, and human accountability. Companies that prioritize these principles will be the ones that effectively leverage AI to deliver superior customer experiences and achieve sustainable growth.
Source: Amatriain, X. (2026, April 14). The AI Trust Gap: Why Travelers Will Continue to Choose Trusted Brands. Expedia Group Blog. https://partner.expediagroup.com/en-us/resources/blog/ai-trust-gap-why-travelers-continue-to-choose-trusted-brands
