ChatGPT Launches 'Study Mode' to Combat Academic Misuse

OpenAI has introduced a new “study mode” feature for its ChatGPT chatbot, designed to promote responsible academic use amid a notable increase in the misuse of artificial intelligence tools within universities.

Accessible via the chatbot’s tools button, the new mode guides users through complex subjects in a step-by-step format, simulating a structured academic lesson. For instance, when asked to explain Bayes’ theorem, a mathematical formula, the chatbot first assesses the user’s current mathematical proficiency and learning objectives before offering tailored guidance.

This initiative directly addresses growing concerns in academic communities. A recent Guardian survey in the UK highlighted the scale of AI misuse, revealing nearly 7,000 confirmed instances of AI-assisted cheating in the 2023-24 academic year. This figure, equating to 5.1 cases per 1,000 students, marks a significant rise from 1.6 cases per 1,000 students in 2022-23. Despite these challenges, OpenAI reports that over a third of college-aged young adults in the US use ChatGPT, with approximately a quarter of their interactions focusing on learning, tutoring, or schoolwork.

The core design principle behind study mode is to prevent the chatbot from simply generating complete essays or direct answers. OpenAI emphasizes that the feature aims to facilitate understanding rather than merely providing solutions. Jayna Devani, OpenAI’s International Education Lead, stated the company’s commitment to preventing student misuse of ChatGPT, positioning the new mode as a step towards fostering constructive academic engagement. She articulated that the initiative represents an effort to demonstrate responsible ways for students to interact with ChatGPT to genuinely support their learning processes, stressing that OpenAI does not endorse the misuse of these tools.

Devani acknowledged that effectively combating academic cheating necessitates a broader “whole industry conversation” to reform assessment methods and establish “very unambiguous guidelines” for responsible AI use. She also noted that students can still bypass this feature to seek direct answers if they choose not to engage with study mode.

Billed as “study and learn” within the chatbot’s options, OpenAI highlights the mode’s particular utility for homework assistance, exam preparation, and grasping new topics. Devani reiterated that the mode encourages user engagement with subjects and problems, aiming to guide users toward solutions rather than providing immediate direct answers. The feature also supports image interaction, allowing students to upload and work through past exam papers with the chatbot’s assistance.

OpenAI developed the tool in collaboration with teachers, scientists, and education experts, but cautioned users about the potential for “inconsistent behavior and mistakes across conversations.”