Concerns about academic honesty in online courses are not new, but the ease with which students can now use artificial intelligence has fundamentally changed the landscape. This presentation is designed to move beyond documenting the problem and instead focus on sharing practical solutions. Over time, we have implemented many strategies commonly used to reduce cheating, including rewriting publisher test-bank questions, randomizing questions and numerical values, imposing time limits on assessments, and redesigning exams to emphasize application rather than recall. While these approaches were once effective, recent advances in AI have made it easier for students to work around traditional safeguards. As the technology evolves, substantially reducing the cost of summoning external assistance, we owe it to our students to continue to evolve such that we provide them with clear motivation to learn in the face of competing demands for their time. Methods to mitigate the use of AI during online assessment will be discussed including institution-provided and market online proctoring systems (e.g., HonorLock), white-type instructions, and time-stamp and grade-metric analysis. These have been reasonably effective in mitigating academic misconduct in online courses. They also have introduced challenges, both to students and instructors worthy of note such as increased technology issues, inequity as some but not all students find a workaround, and the effect of explicit lack of trust measures on the instructor-student relationship. Thus, in addition to enforcement and detection measures such as those noted, we will also introduce additional instructional design and student incentives options (e.g., questions requiring applied reasoning with specialized knowledge unique to the course; short oral exams graded with AI). We will have plenty of content to cover the ten-minute allocation but would prefer to leave more than five minutes for discussion so that other instructors can share their ideas and experiences.