In this innovative practice work-in-progress paper, we hypothesize that in engineering areas such as signal processing, tools such as ChatGPT do not threaten academic integrity in the classroom. We believe that if questions and problems are suitably posed, ChatGPT can assist, but cannot provide solutions. To test this hypothesis, we ask two questions: (a) How can ChatGPT be used to assist students in a signal processing class? and (b) How can the class itself be designed to leverage what ChatGPT has to offer? To answer these questions, we deploy ChatGPT in three different scenarios: (1) In a graduate level course to explore the use cases and limitations of the tool; (2) In summer REU cohorts to study the attitudes of students before and after one-hour workshops; and (3) in undergraduate signal processing courses where students will be exposed to generative AI tools over an entire semester. Surveys and discussions with the students will be analyzed and results will be presented at the conference. With the three separate activities across different time scales and student levels, our results can be used to generate guidelines for instructors to incorporate generative AI tools in their classes.