Artificial Intelligence (AI) proves to be an invaluable tool in numerous areas - from boosting efficiency in production to enhancing customer support, as our last article impressively demonstrated. However, this development also raises questions about responsibility and the legal framework of such systems. A recent case at Air Canada provides interesting insights into this issue, but also shows why we shouldn't overrate this incident.
Recalling our last article, where we discussed the implementation of AI-powered customer support at Klarna, Air Canada, much like Klarna, has adopted an AI-based solution to make its customer service more efficient. However, Air Canada's AI recently made headlines when it mistakenly informed a customer that they were eligible for discounted bereavement fares after the death of an immediate family member, even after purchasing the ticket.
The customer, relying on the AI's information, faced resistance when attempting to claim the discount from Air Canada's customer support, leading to a legal dispute over the responsibility for the AI's misinformation. In an almost humorous twist, the airline tried to argue that the chatbot could be considered a separate legal entity. A Canadian tribunal thought otherwise and ordered the airline to pay about 600 dollars for bereavement refunds and tribunal costs – about half of what the customer had paid for the tickets.
At first glance, this incident might seem like an embarrassment for AI technology, yet it opens up an important perspective: just like traditional software, companies are responsible for the software they deploy, including chatbots. While this case indeed illustrates the risks of AI systems, companies should be careful not to overly correct their assessment of the pros and cons of modern AI systems.
It is likely that Air Canada has made significant savings by automating their customer service. For comparison, in the case of Klarna, which we examined in the last blog post, the AI solution is estimated to lead to a profit increase of 40 million USD in 2024. Thus, the benefits of AI systems significantly outweigh the costs of occasional errors. It is important to recognize that many SMEs have not yet tapped into the potential of AI. They cannot afford to make the mistake now of being unnecessarily worried about deploying AI due to such exaggerated headlines.
Let's consider this incident at Air Canada as a learning opportunity. It underscores the need to carefully implement AI systems while recognizing the incredible benefits they can offer. In a world where AI is increasingly integrated into our daily lives, it is essential that we understand both the technological and legal frameworks that guide its application.
It is clear that AI technology is here to stay, and its potential to improve operational efficiency and customer satisfaction is undeniable. As the Air Canada case shows, the "mishaps" are often not as dramatic as they appear, and provide valuable insights into further development and adjustment of our approaches. It's time to look optimistically into the future and explore the possibilities that AI opens up for businesses of all sizes.