ChatGPT is launching a 'study mode' to encourage responsible academic use of the chatbot, as universities report a sharp rise in AI-assisted cheating. The feature, accessed via the tools button, guides users through complex topics in a step-by-step format similar to a structured lesson.
For example, when asked for help with Bayes' theorem, the chatbot first asks about the user's maths level and goals before offering explanations. This approach aims to avoid simply providing complete essays or answers, with OpenAI stating that the tool 'doesn’t just offer solutions without helping students make sense of them'.
The launch comes amid increasing AI misuse in higher education. A Guardian survey of UK universities found nearly 7,000 proven cases of cheating using AI tools in 2023-24, equivalent to 5.1 per 1,000 students, up from 1.6 per 1,000 the previous year. OpenAI reports that over a third of US college-age adults use ChatGPT, with about a quarter of messages related to learning.
Jayna Devani, OpenAI's international education lead, said the company wanted to show 'there are responsible ways to engage with ChatGPT to actually support a learning process'. She acknowledged that tackling cheating requires 'a whole industry conversation' about changing assessments and clear guidelines on responsible AI use.
The study mode, labelled 'study and learn' in the chatbot's tools, is designed for homework help, exam preparation and learning new topics. It can also interact with uploaded images, allowing students to work through past exam papers. However, OpenAI warned that the tool may exhibit 'inconsistent behaviour and mistakes across conversations'.
Despite the new mode, students can still take shortcuts by ignoring it. Devani emphasised that this is a step towards constructive use, but the responsibility also lies with educators and institutions to adapt assessment methods.
