Will AI Replace Call Centers? The Future of Customer Support

AI-generated image.

AI is poised to revolutionize customer service – but will it spell the end for call centers?

According to Gartner, by 2026 generative AI will have reduced the number of customer-service agents by 20-30%. And in April 2024, Tata Consultancy Services chief K Krithivasan told the Financial Times that AI would result in “minimal” need for call centres in as little as a year.

Bold words that were perhaps a little premature – the telecoms world certainly doesn’t move that fast, even when AI empowered. His timescales might have been a little too aggressive, but is the overall prediction valid?

Reactive, proactive, preventative support

There are three main modes that a call center and other customer support infrastructure can act in – reactive, proactive and preventative. It is this latter mode that K Krithivasan sees as transformational.

“We are in a situation where the technology should be able to predict a call coming and then proactively address the customer’s pain point,” he said.

Modes of Customer Support in Call Centers

• Reactive: Addressing customer issues after they occur.

• Proactive: Anticipating customer calls and addressing issues before customers reach out.

• Preventative: Using AI and data to identify and eliminate potential problems before customers notice them.

We know that high levels of network automation, combined with real-time data insight and powerful combinations of AI and ML capabilities, mean many problems can be anticipated and prevented before they impact customers. Data can reveal that equipment is about to fail before it does – providing a window of opportunity for engineers to replace it proactively; networks can detect localised loading and adapt their resources to prevent problems associated with congestion. Historic and contextual customer data can be used to anticipate customer need – ensuring that patches are applied, firmware is upgraded, CPE and handsets are not causing problems.

We can even use network data to provide a rich picture of what customers are doing in order to design new, appealing products informed by customer behaviour and then market these to those customers most likely to buy them, ensuring customers have the right products to support what they’re doing at any given moment.

Impact on call center jobs

All of this is revolutionary. But at stake is a huge industry that provides millions of well-paid customer service jobs. Nearly 17 million globally – with 3 million in the US and over 800,000 in the UK – with the Philippines contributing 1.5 million agents and retaining a 16% share of the global outsourcing market. Other significant destinations for call center outsourcing include India, Brazil, Colombia, Poland and South Africa.

Global call center job market snapshot 

• 17 million call center agents globally.

• 3 million agents in the U.S.

• 800,000+ agents in the U.K.

• 1.5 million agents in the Philippines (16% of global outsourcing market share).

• Emerging markets: India, Brazil, Colombia, Poland, South Africa.

Why human agents remain essential

But bullish as Krithivasan is about the impact of AI – at the time of writing, he reported to the Financial Times that his company’s generative AI projects had doubled quarter over quarter and they were enjoying a record order book for FY24 – even he doesn’t believe that the technology will lead to an overall reduction in jobs, just a change in what people are doing.

“The world is going to need more and more people, not fewer people, in terms of technology talent,” he said.

After all, AI is a tool and someone needs to train the models, oversee them and set the business objectives. While some jobs will no longer be needed, these will tend to be those involving a lot of repetitive tasks – such as writing up an interaction with a customer in a call centre. By taking over such mundane tasks, AI will free up human employee’s to do more of the things that they are uniquely good at – such as creative problem solving and empathy.

Strategies for dealing with the shift to AI

Overall, the World Economic Forum predicts that AI will displace 92 million roles, while creating 170 million new jobs. They argue that bridging the skill gap is one of the biggest challenges facing employers, although they say 9 out of 10 employees are capable of being trained, upskilled or redeployed. Respondents to their Future of Jobs Survey, revealed a mixture of strategies for dealing with the shift to AI – with 85% planning to upskill their existing workforce, 70% planning to hire staff with new skills, 40% planning to reduce staff with obsolete skills, and 50% planning to transition staff from declining to growing roles.

In the customer support world, humans will never be entirely eliminated, although their roles and skills will change. Models need solutions and data to learn from, and this relies on human creativity – in other words, human agents and engineers solving new problems in real-time. AI models can learn from these solutions, but can’t creatively solve entirely new problems – at least yet. Likewise, a proportion of problems will be uncommon or even unique, meaning it’s unlikely to be cost-effective to train an AI model to solve them.  

Two main reasons why humans will always be required

After all, AI is a tool and someone needs to train the models, oversee them and set the business objectives. But even in a perfect world, the call center is unlikely to ever be entirely eliminated.

There are two main reasons why humans will always be required.

• Firstly, models need solutions and data to learn from, and this relies on human creativity – human agents and engineers solving new problems in real-time. AI models can learn from these solutions, but can’t creatively solve entirely new problems – at least yet. Likewise, a proportion of problems will be uncommon or even unique, meaning it’s unlikely to be cost-effective to train an AI model to solve these if they’re very low volume or unlikely to occur again.  

  
  

• And then there’s the issue of empathy. Sure, AI can recognise emotional cues, analyse interactions to detect anger or distress, flag up warnings in the co-pilot, and provided real-time adapted scripts based on human psychology to assist the agent in dealing with the customer. All very helpful. But is it effective?
  

  
  

Balancing AI efficiency with human empathy

Writing in The Economist, Sophie Elmhirst recently related how a friend had recently been reduced to tears while on the phone to her broadband company. “The agent tried to make her feel better. ‘It’s only broadband!’ But it’s not only broadband, is it? It’s the distinct sensation of life slipping between your fingers, the feeling that you are lost in a dehumanised process enacted by a giant machine which doesn’t care about either its customers or its employees. It is automated hell.”

This feeling isn’t unique to Elmhirst, a year after predicting substantial job losses in the call center market, Gartner revealed that 64% of customers said they would prefer companies not to use AI in for customer service, and 53% of customers would consider switching to a competitor if they found out a company was using AI for customer service.

“Many customers fear that GenAI will simply become another obstacle between them and an agent,”

explained Keith McIntosh, Senior Principal, Research, in the Gartner Customer Service & Support practice.

In fact, research by the Harvard Business School has shown that 30-50 per cent of people are willing to wait hours for a human to respond to their problem, rather than receive an instantly generated AI response. Researchers concluded that customers place a high value on human empathy and connection, and in certain circumstances are willing to sacrifice time in order to get it.

Customer preferences around AI-driven support

• 64% prefer companies not to use AI for customer service.

• 53% would consider switching providers due to AI-based customer service interactions.

• 30–50% are willing to wait significantly longer for a human agent due to perceived empathy and connection.

The evolving role of AI in Customer Support

This suggests that AI’s main roles are supportive and preventative – making interactions more efficient and freeing up agents to make outbound calls to check in on customers and provide genuine empathy when required. All of which supports the notion that the call centre will not die, but transform into an interaction and empathy hub, sitting at the heart of the relationship between customer and the telecom service provider.

Because, as the CEO of The Empathy Business, Belinda Parmar, told the Financial Times more recently, the goal of AI should be to “enhance human connection, not replace it”. As for call center jobs, despite the onslaught of AI, the industry is still growing with about 70,000 extra agents created globally in 2024 with 110 sites created or expanded.

AI's evolving role in Customer Support

• AI's role is primarily supportive and preventative, enhancing efficiency.

• Frees up human agents to engage in proactive, empathetic customer interactions.

• Transforms call centers into interaction and empathy hubs, strengthening customer relationships rather than replacing them.