The Wall Street Journal headline – “How a Chatbot Went Rogue.” – is scary, as is the story that follows. A chatbot accessible through the website of the National Eating Disorder Association’s website, with output determined in part by an AI system called Tessa, dispensed dietary advice widely viewed as inappropriate to some members of a psychologically vulnerable audience, raising fears of serious mental trauma or resultant physical harm. According to the bot’s maker, Tessa was unexpectedly drawing on open-ended data sources, rather than only an intended, pre-approved set of responses.

You’ve probably seen a few scary headlines about how the output from some AI systems can be offensive or off-topic. And it’s true: An open-ended generative AI platform–one trained on billions of data points and relatively unconstrained in its output–can come up with results that are interesting and unexpected, but also potentially harmful.

More often than not, the two terms are used interchangeably, a trend that has left the marketing and business communities confused. While both of these tools are used in similar business procedures, the extent to which...