Bot detection models may return a high rate of accuracy, but that’s due to a critical limitation in the data used to train them, a new study finds.

Before it’s in the boardroom, it’s on the syllabus at MIT Sloan — including a new course on pursuing a meaningful life.