It’s Time to Ensure that all Chatbots are Ethical

December 17, 2019

It often seems like chatbots are everywhere – wanting to strike up conversations with you when you’re on a website or calling into customer service. But those are just the chatbot conversations you know about. Chances are you’re communicating with chatbots in other instances and you’re not even aware of it.

One of the pressing issues the industry faces is in how we are using chatbots and other forms of Artificial Intelligence (AI) in a responsible and transparent manner. In fact, I was interviewed on this topic recently by the technology editor of Strategic Finance, and his story appeared in the publication’s December issued. In the article, I shared the importance of ensuring that we are developing and deploying ethical AI algorithms.

As I explained, first and foremost, transparency must be the watchword. In every chatbot interaction with customers, companies should inform them that they are interacting with software and not people. Organizations should be sensitive to when it is appropriate to use chatbots and when it’s not. No one wants to hear bad news from a doctor’s office, for example, delivered by technology. We also need to make sure that organizations are culturally sensitive and can communicate appropriately with diverse audiences. One of the ways we do that at Wovenware is to build every app in the language in which it will be deployed, using native speakers to avoid inadvertent mistranslations or misleading idioms.

Additionally, since they are AI apps, chatbots are continually learning. Companies must make sure they are not learning the wrong things from interactions with humans, like when Microsoft’s Tay chatbot began spewing antisemitic and other racist rants. To avoid this, companies must continually monitor their chatbots and provide ongoing corrections, as needed.

Chatbot technology has matured quickly and will continue to evolve, so it’s critical that we are controlling it, rather than being controlled by it. We must address this now, and consider regulations to set the parameters of acceptable chatbot behavior, as the technology continues to grow in popularity.

Leave a Reply

  • (will not be published)