Home > Columns > CRM Columns

Addressing the AI Elephant in the Room: Trust

CrmXchange

Presented By: CrmXchange



Contributed article by Seth Johnson, CTO at Cyara   

Unless you’ve been living under a very large rock, you’ve most likely interacted with Artificial Intelligence (AI) – whether in a personal setting engaging with bots like Alexa or Siri, or in a business setting using apps like ChatGPT or Claude, or less obvious automation use cases to improve workflows. In the contact center space, the proliferation of agentic AI is especially prevalent as the use cases are obvious, and the benefits to enterprises are numerous – from improved customer experience, to reduced agent turnover and overall cost savings.   

While enterprises are quick to adopt AI agents to reap those aforementioned benefits, their haste to do so may actually be costing them more in the long run. That’s not to say AI can’t cut costs – it can, and it does. The challenge is that many leaders mandate AI agent deployments without implementing the proper training and guardrails, resulting in inefficiencies, hallucinations, and clunky customer experiences (CX) that damage brand reputation and loyalty, leading to lost revenue. Overall, it comes down to trust between brands and their customers.   

In this piece, we’ll explore the top CX mistakes brands make, according to customers, and address steps enterprises and brands can take to ensure AI and AI agents are enablers of trust, not detractors.   

Top CX Dealbreakers for Customers 

“Speak to agent” is an all-too-familiar trope of customer service interactions, and data that Cyara recently commissioned confirms it – the number one CX dealbreaker for consumers is not being able to reach a live human. Further, 73% believe human agents resolve issues faster than AI bots. However, expectations are also higher, with 87% saying they have higher issue resolution expectations for humans than for either chat or voice bots.   

This is both a trust issue, and a perception one. Even with advances in agentic AI, it’s difficult to replicate true human-to-human interactions and the inherent trust that comes along with that, but there’s a misconception that humans tend to resolve issues the fastest or most comprehensively. When it comes to issue resolution, data shows that technologies like agentic AI can actually solve many issues more efficiently and with shorter wait times. Approached properly with testing and monitoring, conversational AI can lead to an 80% boost in self-service interactions that don’t depend on a human to resolve the issue.   

Additional CX dealbreakers from the survey include long hold times and slow responses, having to repeat themselves across multiple agents or channels, and a lack of resolution or unclear next steps after an issue. Each of these can cause strong negative customer reactions, leading to frustrations and loss of loyalty. They are also strong use cases for implementing AI, which can easily cut down on wait times and mitigate the dreaded information repeat dilemma while guiding customers to clear next steps.   

Finding the Balance Between AI Automation and Human Empathy 

Listen, I am not the first to say that it’s not humans vs. AI, it’s humans and AI. But it does bear repeating. Keeping a human in the loop will be a point of distinction, particularly for certain demographics and use cases. For example, 81% of Baby Boomers favor human agents for resolving issues faster, while Gen Z has a more positive sentiment toward AI bots, with 21% saying that human agents and AI bots resolve issues in a similar timeframe. Looking at use cases, 65% wouldn’t trust bots to handle financial/security transactions and 53% are skeptical about bots handling sensitive healthcare data. That’s not to say that enterprises should abandon AI for these use cases altogether, but rather they need to be careful and strategic about their deployments, and be transparent with customers about AI’s use – while always ensuring they have access to a human.   

There’s also plenty of opportunity for AI to improve the experience in the background for better agent experiences and outcomes. This helps keep a human in the loop, instilling strong trust and ensuring empathy is front-and-center, while still reaping the benefits of AI’s intelligence and automation capabilities.   

Modernizing CX and the Case for AI Accountability 

Modern CX requires modern solutions, but it doesn’t mean AI without guardrails is the catch-all to resolve all of CX’s challenges. As enterprises look to not only experiment but fully deploy tech such as agentic AI, it must be thoughtfully designed and deployed. The key to AI’s success lies in its ability to resolve issues and build trust, and to do that, leaders must ensure proper guardrails, testing and monitoring. You wouldn’t buy a car without seatbelts or one that didn’t undergo rigorous safety inspections, and you shouldn’t implement AI without assurances that it will work how it’s supposed to and lead to stronger CX outcomes. At the end of the day, if your AI agent hallucinates, makes a mistake or gives improper guidance, you can blame a rogue agent all you want, but it’s your brand that will suffer reputational and revenue losses. CX assurance is not a nice-to-have for brand reputation – it’s a critical must-have for brand survival.   

The real benefits of AI implementations in the contact center are too great to ignore – but so, too, are the risks. The onus is on leaders to deploy AI agents responsibly to build trust with customers. Their livelihood and brand reputations depend on it.