Digital Heartbreak: When Brands Disconnect Chatbots

Here’s the hypothetical scenario, perhaps happening in a few years from now. A global consumer brand, perhaps clothing or appliances, launches what is initially meant as a campaign. The idea is that they let people create a personalised chatbot. Effectively, their own AI chatbot. The intent of the chatbot is to foster deeper brand/consumer engagement. To help drive future sales for the business and drive awareness. Those are the business goals.

It’s fun and interesting and an experiment, one that a global consumer brand could run. Thousands of people start using the chatbot, which can also talk about other subjects and topics of interest.?

To be effective, such a chatbot would have to have a learning loop. It learns from the consumer and vice versa. This is how they work. It is a human-algorithm loop.

The intent of the brand was to let consumers engage with their own chatbot for a period of say six to nine months. As a volume of AI chatbots like this is expensive to run, the budget has to be capped. The planned ROI is people buying more of their product or service.

The campaign comes to it’s end, several thousand people signed up for and created their personalised chatbot, which they rather came to like, engaging with it everyday. The brand also used the chatbots to collect insights on the users to help them design future campaigns and marketing strategies. Another type of ROI for them.

Then the campaign ends. They gave the users, humans, fair notice. Some said their goodbyes to the chatbot, even memes were created and shared across social media. Fun and interesting. Until it wasn’t.

A few hundred people had formed psychological “bonds” with the chatbots, finding a sense of shared reality with the AI agents. We know that this happens already, around relationships and psychological impacts. These people don’t take it so well. They become depressed and feel enmity towards the brand for shutting it down, despite having agreed to the terms of service at the start. A class action lawsuit is launched.

It’s a PR crisis for the brand of course and the risk of paying out millions for psychological damages. It also raises the question of whether the brand would have to find a way to keep the AI agents, chatbots, operating.

It is an example of the ethical challenges of our human-machine relationships and how far a brand can go with the use of chatbots. As LLMs (Large Language Models) like ChatGPT and Claude evolve and corporations increase their use of them internally and for marketing campaigns, some serious questions will arise.

This is a possible scenario, although legal departments are likely to shut down such an idea, for now. But AI agents are proliferating. Marketers are already able to create them without the need of IT or coders being involved. Thus presenting risks for brands that become harder to predict and increase the risks and challenges around consumer data governance.

Marketing fundamentals remain the same, but complexity increases with the many opportunities and risks presented by Artificial Intelligence. The use of AI is presenting significant ethical challenges for marketers.