Contact Centres Are Not Ready for Generative AI

Contact centres are the lifeblood of many businesses. They provide an essential service for customers, driving satisfaction and loyalty. With the rise of artificial intelligence and machine learning, many contact centres are leaning towards using these technologies to increase efficiency and improve customer service.

The challenge is that, while the potential benefits of automation are enticing, the current state of technology and infrastructure falls short in several critical areas. In this article, we explore why contact centres are not yet ready to fully embrace generative AI.

Lack of High-Quality Training Data

One of the most important aspects of AI is its ability to learn from data. In order for generative AI systems such as ChatGPT to provide accurate insights and responses, they need access to high-quality data sets. This is especially true for contact centres, where customer interactions can be unpredictable and complex. Unfortunately, many contact centres lack the adequate volumes and quality of data necessary to effectively train AI systems, and the solution to this isn’t totally straightforward. Why? There are three key factors coming into play here.

1. Building, training and deploying AI models is a time-consuming undertaking. Businesses interviewed by the IDC (International Data Corporation) in the US stated that “much of their AI development time was just spent on data preparation alone”. This process involves collecting considerable amounts of quality data, cleaning and labelling it before being able to use it for training. This alone puts a huge strain on resources, and can hinder the deployment of AI at scale.
2. Privacy, security and GDPR compliance must be respected. This adds another hurdle to gathering the training data needed, and can lead to incomplete or inconsistent data sets. This barrier, combined with the sheer scale of data needed to adequately train AI systems, is so significant that many businesses have chosen to outsource their contact centre operations as a cheaper alternative.
3. Bias, both human and algorithmic, can often creep into data sets. If there is bias in the data used to train AI systems, which can inadvertently happen when collecting wide-scale datasets, it can lead to customers receiving inaccurate or irrelevant advice, and worse yet, discriminatory responses.

In addition to the above, it’s also important to consider that the training data used to deploy an AI system may become outdated or irrelevant. The customer service landscape is constantly changing, and AI systems need to be able to keep up with these changes to remain useful.

‘AI models are not capable of fully understanding complex customer interactions and providing the same level of personalised service that a human can.’ Jason Roos, CEO of Cirrus

Complex and Dynamic Customer Interactions

Customer service representatives deal with a wide range of customer inquiries, from mundane to complex. Providing accurate answers requires a deep understanding of both context and customer intent. While AI can generate human-like responses, it can still struggle to understand the nuances and intricacies of human interactions. In addition, the non-linear structure of natural conversations can confuse the system and lead to inaccurate responses.

In contact centres, agents are trained to manage the challenges of customer conversations. They have the ability to interpret customer sentiment, make sense of ambiguous or conflicting information, and deal with difficult customers. AI, on the other hand, lacks this level of interactivity and emotional intelligence. As Jason Roos, CEO of Cirrus, states: “AI models are not capable of fully understanding complex customer interactions and providing the same level of personalised service that a human can. The limitations of generative AI are such that they will never be able to compete with the power of human intelligence. Well-trained agents remain core to any contact centre business.”

Legal and Ethical Considerations

AI is a powerful technology with far-reaching implications for our society. It’s no surprise that governments are starting to create laws and regulations around the use of AI. For example, the AI Act proposed by the European Union sets out principles around safety, responsibility and transparency when using AI. Once approved, this will be the world’s first legal parameter on AI, and it likely won’t stop there.

When it comes to customer service, there are ethical considerations that must be taken into account. As AI models store considerable amounts of personal data, this raises concerns around misuse of sensitive information, data breaches and unauthorised access. To address these concerns, robust safeguards, transparency, and ethical frameworks must be put in place when implementing generative AI in contact centres. AI should also never replace human judgment or discretion, as decisions are best made when taking into account the complexities of each customer’s situation along with ethical considerations.

Trust and Human Touch

We may be becoming accustomed to digital customer service, but customers still prefer to interact with human beings. A recent study found that 70% of adults trust people more than technology. While AI may be able to provide accurate responses in many cases, customers still prefer to interact with real people who can provide personalised and empathetic support.

The impersonal nature of AI can make customers feel like they’re not being listened to or taken seriously, which can lead to customer dissatisfaction and attrition. In contact centres specifically, trust and human touch are essential components of creating a positive customer experience. AI should be leveraged to augment agents, not replace them.

Continuous Learning and Adaptability

Contact centres are constantly evolving, and AI systems must be able to keep up with the ever-changing landscape. As customer demands evolve, so too should AI models. They need to continuously learn, adapt, and improve their responses based on real-time feedback and emerging trends.

To do this, AI systems must have the ability to capture and process vast amounts of data in real time – something current technology is yet to achieve. While the race to achieve this is underway, it’s still highly doubtful that technology will ever be able to match the agility of human agents. It will, however, be increasingly able to assist them in their day-to-day tasks, so that customer interactions are more efficient and effective. The key is striking the most rewarding balance between the best of human and machine capabilities.

Conclusion

With the continuing development of AI, contact centres are facing disruption like never before. To stay ahead of the curve, businesses must ensure that AI is properly implemented and used to augment their customer service operations – not replace them. It’s clear that contact centre agents are still an integral part of the customer experience, especially as the technology doesn’t yet match the complexity and agility of human interaction.

Ultimately, what’s important is that contact centres find the right balance between AI and human agents to provide the best possible customer service. With this approach, contact centres can ensure that they continue to not just meet, but exceed customer expectations.

At Cirrus, we have developed best-in-class AI-powered solutions to help contact centres to leverage the power of AI. With our intuitive and powerful AI tools powered by ChatGPT, contact centre agents can automate mundane tasks, and increase efficiency and accuracy to elevate the customer experience. Get in touch today to learn about how we can optimise your customer service operations.