OpenAI Adds Interactive Quizzes to ChatGPT on Web and Mobile
Consumer and Edu users can answer quiz questions inside a chat, but OpenAI has not published assessment-quality or learning-outcome evidence.
OpenAI added interactive quizzes to ChatGPT on August 14, 2026. People can ask ChatGPT to quiz them on a topic and answer questions inside the same conversation, shifting the chat from giving an explanation to supporting a round of active practice.
The change is available across all consumer ChatGPT plans and ChatGPT Edu on web and mobile. That reach matters for learners and educators because the interaction sits in the existing chat surface, but OpenAI has not published evidence that it improves learning outcomes or produces reliable assessments.
How the quiz interaction works
OpenAI describes the feature as in-conversation practice: a user names a topic, asks to be quizzed and submits answers directly in the chat. The announcement does not specify a fixed subject list, question format, scoring method or adaptive-difficulty system. It also does not say whether quizzes can be saved, shared, assigned or reviewed through educator controls.
Those omissions keep the supported claim narrow. This is a product-interface release, not evidence of a new teaching model, a validated assessment system or a replacement for an educator's judgment. The source establishes that the quiz flow is available; it does not establish how accurate, challenging or instructionally useful any generated quiz will be.
Availability spans consumer and Edu plans
The official note names all consumer ChatGPT plans and Edu plans, with access on both web and mobile. It does not provide a region-by-region rollout schedule or a separate availability statement for other workspace plans. Users should therefore treat the published plan and platform list as the confirmed scope, without assuming additional eligibility.
For students, the practical change is fewer setup steps between asking for an explanation and checking recall. For educators, it creates another possible practice surface inside ChatGPT Edu. Neither implication proves that the generated questions align with a curriculum, measure mastery or handle misconceptions consistently.
Evidence quality and limits
The evidence is one official OpenAI changelog entry. It is authoritative for the fact that OpenAI released the feature and for the named availability, but it is not independent testing. OpenAI supplies no public results on quiz accuracy, learning gains, accessibility, age-specific safeguards, educator administration or performance across subjects.
Because generated questions and feedback can shape what a learner remembers, the next useful evidence would be documentation on quiz construction and controls, followed by independent studies comparing the experience with established practice methods. Until then, the feature is best understood as a broadly available interaction option whose educational quality still needs measurement.
Status
Confirmed. OpenAI's official release notes document the August 14 change. Internal confidence is medium because the evidence comes from a single first-party changelog and includes no independent evaluation.
Sources
Update note: Last reviewed 2026-08-17. We will revise this post if OpenAI publishes fuller availability, controls or evaluation evidence.
Sources
- OpenAI — ChatGPT release notes — official
Drafted with AI assistance from source briefs; reviewed for citation completeness and label accuracy.