A study conducted by Cornell University highlights the growing trend of generative AI misuse among college students, revealing that 37% of students use AI tools like ChatGPT for assignments, with 9% admitting to using them to cheat. The research, which analyzed responses from over 95,000 students across 20 public universities in the U.S., indicates significant variations in usage based on academic discipline and demographics. For instance, 62% of computer science students reported regular AI use, compared to only 24% in the arts. Additionally, the study found that 45% of male students used AI regularly, versus 33% of female students, suggesting potential equity gaps in access to technology. The researchers employed a list randomization technique to estimate cheating rates, finding that 26% of daily AI users had cheated, compared to 7% of monthly users. To combat these challenges, the authors recommend assessment reforms, including controlled testing environments, clearer guidelines for AI use, and adapting assessments to incorporate AI in a manner that reflects professional skills. They caution that without careful implementation, these reforms could worsen existing educational disparities.