Taking ample space on today's newspapers, blogs and social media, artificial intelligence (AI) chatbots are rapidly conquering society. First, ChatGPT (short for Chat Generative Pre-Trained Transformer) by OpenAI launched in November 2022, and was met with widespread excitement, as well as serious concern. As the New York Times wrote, "not since the iPhone has the belief that a new technology could change the industry run so deep."

A technology that revealed itself to be so transformative means that numerous companies started to seek ways to benefit from it, working on creating their own AI-powered chatbots.

In February 2023, Microsoft announced a new version of their search engine Bing, in which users can search via conversational prompts, powered by the same technology as ChatGPT. Around the same time, Google announced that they are working on their own AI-powered chatbot, Bard, likely in response to the immense noise made around the public nature of OpenAI's ChatGPT.

So what is ChatGPT? And how do you use ChatGPT?

ChatGPT, Bard, Bing, and other AI agents, are the latest innovation in "large language model tools", a technology that doesn't 'think' in the same way that humans do, but rather has an exceptionally large amount of data and computing techniques that allow it to make predictions to string words and sentences together, and anticipate where the sentence can go. Unlike their predecessors, these tools are particularly good at not only tapping into the extensive vocabularies and information at their disposal, but also at utilising context to mimic human speech patterns and understand questions and prompts.

ChatGPT is conversational, meaning that users can type prompts or questions as if sending a message in a chat conversation. Then, unlike search platforms of the past, which would respond with a list of websites where users can seek out the information they are looking for, ChatGPT will respond with full sentences to the users' inquiry or request. As of now, unless explicitly prompted, the tool does not cite its sources.

Use cases for ChatGPT in customer service

In the last several months there have been a boom in articles and think-pieces musing about how ChatGPT and other AI chatbots may be deployed for use in business, academia, integrated logistics, and domestic life. From writing emails, essays, and video scripts, to providing travel tips and recipes; ChatGPT is seemingly being put to use in wide-ranging ways.

So, how might technology like ChatGPT be utilised in customer service and care?

  1. Providing personalised support: Since ChatGPT and other AI chatbots use conversational language, they can be deployed to identify common customer concerns and provide solutions accordingly.
  2. Enhancing engagement through 24/7 availability: Instead of having to always have a social media or customer care manager online, AI chatbots can respond to messages in real-time, allowing for quick response times and better engagement across channels.
  3. Improving experience: By collecting and analysing customer feedback and questions, AI chatbots like ChatGPT can find patterns on how products and services could be improved for better customer experience.
  4. Providing recommendations: Utilising data on past purchases and shopping habits, AI chatbots can provide customers with product recommendations, increasing sales and boosting loyalty.

The importance of the human touch in customer centricity

There is a “dark side” to AI too. Despite the 'human' way that tools like ChatGPT operate these tools are far from human, and the limitations that come from their use should be remembered. There are innumerable examples of AI chatbots misbehaving and providing false information like citing fake sources, making up answers, or even engaging in inappropriate behaviour.

Especially in contexts where businesses aim to utilise ChatGPT in a customer care function, a cautious approach to implementation should be the path forward. Some ways that technology like ChatGPT could be detrimental in customer support functions include:

  1. Privacy concerns: As these models use the data and conversations that are input into them to further their learning, confidential or proprietary data that is input into the chatbot is no longer secure.
  2. Limited understanding: ChatGPT 3 has been trained on data up until 2021, so when it comes to more timely, newsworthy, or modern inquiries; the technology is limited.
  3. Lack of empathy: A huge concern when it comes to customer centricity is that the chatbots have less capacity for empathy, so while deploying the bots to answer rote questions can be super helpful, there may be moments when responses are less human than is called for.



Especially in contexts where businesses aim to utilise ChatGPT in a customer care function, a cautious approach to implementation should be the path forward. While it is important for companies to be technologically-literate, and learn new digital capabilities as they are developed, blind adoption should never be the norm. Ajay Agarwal, a professor at the University of Toronto focused on AI, recently explained his concerns about utilising ChatGPT for customer service. “ChatGPT is very good for coming up with new things that don't follow a predefined script. It's great for being creative. It's great for asking a range of questions, but you can never count on the answer…and companies wouldn't ever want that as how they are responding in their customer service office.”

Partnering with businesses that have the same tech plan as you

Because this type of technology is so new and comes with numerous risks involving not only data security but also validity, it makes sense for users to tread lightly. Unfortunately, it is common to see companies that are wholeheartedly embracing AI models, without using necessary caution to protect their customers therefore putting their data and reputation at risk.

Businesses that try to balance between AI deployment and personalised (human) customer experience, are likely to be more resilient whenever this technology does misbehave. When it comes to logistics then, it makes sense for such businesses to partner with likeminded companies, especially where their supply chain is concerned. Working with an integrated logistics provider that innovates and digitalises is important for future-proof supply chain management and transportation, but it is important to choose a partner that also understands the limitations and risks of AI, remaining truly customer centric to ensure resiliency and sustainable growth.