Coursera's AI-Graded Peer Reviews for Professional Certificate Courses

Coursera piloted an AI grading feature for the open-ended, text-based peer-review assignments used in many of its Professional Certificate courses (co-branded with partners such as Google, IBM and Meta). Instructors supply a grading rubric, and the AI model scores learner submissions against it, supplementing or replacing peer and teaching-assistant grading; the feature is limited to text submissions and is excluded from Coursera's for-credit degree programs.
In its own beta report (published September 2024), Coursera states the system graded roughly 300,000 submissions, cut average grading time from about 15 hours (human graders) to about 1 minute, delivered roughly 45 times more written feedback per response (averaging 326 characters), and that 90% of learners rated the AI feedback positively. Course completions within a day of a peer review rose 16.7% after the feature's introduction. However, Coursera's own data also shows first-attempt pass rates fell from 88% to 72%, with AI-assigned grades averaging about 3% lower than prior peer grades; only 7% of learners opted to switch back to human grading.
Independent reporting from Class Central raised specific accuracy concerns with the early beta: it found the AI grader sometimes appeared to score based on the tone or "sentiment" of a submission rather than strict rubric alignment, produced inconsistent scores across repeated grading of the same text, and required correction from human graders — particularly those who had completed the course themselves. Duke University's continuing-education arm ran its own opt-in pilot inviting learners to compare AI and human grading directly, indicating that even adopting institutions treated the accuracy question as still open rather than settled.

Read the full analysis: https://blog.coursera.org/ai-grading-in-peer-reviews-enhancing-courseras-learning-experience-with-faster-high-quality-feedback/