Introducing study mode in ChatGPT

OpenAI rolled out something it calls study mode inside ChatGPT this week, and the pitch is simple: stop dumping finished answers on students and start walking them through the work instead. Ask a calculus question in study mode and the bot won't just spit out the derivative. It asks what you've tried, breaks the problem into smaller pieces, and checks whether you actually understood the last step before moving to the next one.

This is OpenAI trying to answer a complaint that's followed ChatGPT since the day it launched in classrooms — that it turns homework into copy-paste theater. Teachers have spent nearly three years watching kids get correct answers with zero comprehension behind them. Study mode flips the default interaction from answer-first to process-first, using prompting techniques that look a lot like a patient tutor: guiding questions, scaffolded hints, and feedback loops that force the student to do the thinking.

What's notable is that this isn't a separate app or a paywalled feature bolted on top. It lives right inside the regular ChatGPT interface, just a mode you switch into, which means the barrier to trying it is basically zero. Students who'd otherwise ask for the final answer now get nudged toward showing their own reasoning first.

Whether this actually changes behavior at scale is the real question. A tool that can teach patiently is only useful if people choose to use it that way, and plenty of students will still hunt for the fastest path to a finished essay or solved equation. OpenAI is betting that making the slower, more educational path the default — rather than an opt-in extra — will nudge enough people toward it to matter.