AI ethics: AI etiquette

- If every company had an unofficial generative AI ambassador, it would be Rahim. He's not only an early adopter of technology, he actively pushes the boundaries of digital automation tools, and it's paid off. He was awarded top salesperson last year. In sales, it's all about delivering on quotas, and Rahim makes the most of his time to deliver results. That's why he's leaning into using generative AI along with other automation tools for pretty much every task. The first to go were annoying administrative things like expense reports. He then started using it to draft prospecting emails and has set up a series of follow-up steps and communication streams with automated scheduling tools. He's also put existing customers on a rotation of regular communications using generative AI. In order to get the data to make all of this work, Rahim's AI transcription bot follows him to every meeting, recording every interaction, which is then summarized as notes and automatically uploaded into every company's CRM system. Recently, Rahim has taken things one step further. He has cloned his voice and trained his own AI agent. This digital version of Rahim has represented him at meetings with prospects, customers, and even at every company's staff meetings. Rahim can literally be in multiple places at once. While Rahim has not disclosed that he's using an AI agent, some have suspected that they are interacting with AI and not with the real Rahim. Others, however, have been oblivious, thinking the interaction was with the real Rahim. Remember our story at the beginning of the course about the early days of cars? Rahim's full-on embrace of AI without really thinking about the consequences of its use is reminiscent of being able to drive a car without a license and without any rules. Sooner or later, the impacts will be felt. Let's touch on a few of the issues. Firstly, Rahim needs to know that not everyone is as comfortable with AI as he is. He's privileged his desire for efficiency over the comfort and consent of others. It's reminiscent of our earlier discussion about the prudential, doing what's in our own best interests and how that can sometimes be misaligned with ethical actions, which typically involves considering others. For example, Rahim doesn't ask if he may transcribe a meeting using an AI tool. Instead, he has set his bot to automatically show up, assuming that it's okay with everyone. This illustrates a lack of respect for others. Rahim is now sending his digital doppelganger to meetings, and because the bot sounds like him, he's actively deceiving people. Once again, the desire to be productive and efficient through the use of automation is prioritized in ways that are both creepy and duplicitous. One of the basic tenets of responsible AI is disclosure, letting others know that they are interacting with an AI system. Rahim's choices are putting both his and every company's reputation at risk. However, in the short term, he's making a lot of sales and driving revenue, and that makes it harder for Rahim to course correct. Finally, because it's so easy to automatically follow up, Rahim is bombarding prospects and customers with a series of poorly written communications. Instead of being thoughtful, he's churning through leads and creating ill will. This will ultimately hurt customer relationships in the long run. Rahim's behavior raises the important issue of AI etiquette. While the specifics may vary with context, generally speaking, disclosing the use of AI is a good starting point. Asking for consent to collect data by recording a conversation is another normative best practice. Rahim may also wish to include a disclosure notice on automated email correspondence, such as written generative AI, as yet another point of transparency. In addition, some of Rahim's actions are not just unethical. They might also be illegal, at least in the United States. The Federal Communications Commission, or FCC, recently implemented new rules to regulate AI-enabled robocalls in order to protect consumers. Rahim's AI agent might fall into this category. This is an example of an existing law related to robocalls being updated to address new concerns related to AI, and it illustrates how important it is to seek legal guidance when implementing AI tools. Imagine that you are Rahim's manager, and you've just learned about how he's using AI. What would you say to him to encourage his ambitions while also ensuring the ethical use of AI? Here's a hint. Refer back to our framework for making ethical choices. AI etiquette is about using AI in ways that best uphold human values and respect human dignity.