AI agents are completing entire online courses for students, prompting platforms and universities to adopt new detection methods after the Coursera-Udemy merger.
AI-driven software is being used to complete entire MOOCs on behalf of enrolled learners.The practice has intensified after the Coursera-Udemy merger in May 2026 and is prompting institutions to revise integrity policies.
The New York Times reported that generative-AI agents are completing full online courses for students on platforms such as Coursera, edX, and Udemy, effectively substituting human effort with automated outputs [1]. The phenomenon has been documented since the widespread release of large-language models in late 2022, and its visibility increased after the Coursera-Udemy merger announced in May 2026 [2].
Students, online-learning platforms, and universities are the primary actors in the emerging cheating ecosystem. Platforms are deploying AI-powered proctoring tools, while universities are revising honor-code policies to address automated misconduct [2][4]. The cheating process involves prompting models such as ChatGPT, Claude, and Gemini to generate essays, solve problem sets, and produce code, which are then submitted as the learner’s original work [4][1].
Timeline and Geographic Reach
Generative-AI tools became broadly accessible to the public in late 2022, leading to a measurable rise in academic misconduct that includes AI-generated submissions [4]. By early 2024, several universities reported spikes in plagiarism-detection alerts that correlated with the adoption of large-language models [3]. The issue gained further prominence in May 2026 when Coursera completed its merger with Udemy, creating a combined marketplace that serves more than 250 million learners worldwide [2]. The merger prompted both companies to announce new detection protocols aimed at mitigating AI-assisted cheating across their joint catalog of courses.
The cheating incidents are not confined to a single region; institutions in North America, Europe, Asia, and Africa have reported cases of AI-generated coursework passing through automated grading systems [1][3]. MOOC platforms that host courses from partner universities therefore face a global integrity challenge that spans multiple time zones and regulatory environments.
Higher-education institutions that license courses on these platforms are also participants, as they must align institutional honor codes with the technical capabilities of the platforms [4].
Zoho founder Sridhar Vembu and Skyroot Aerospace CEO urged Indian students to forgo conventional degrees and focus on entrepreneurship, citing U.S. examples.
Students are the direct users of AI agents, employing them to complete assessments, write essays, and generate code without personal involvement [1]. Online-learning providers—including Coursera, edX, and Udemy—are responsible for delivering course content, monitoring learner activity, and enforcing integrity standards [2]. Higher-education institutions that license courses on these platforms are also participants, as they must align institutional honor codes with the technical capabilities of the platforms [4].
Coursera has implemented AI-driven proctoring that monitors webcam feeds, microphone input, and keystroke dynamics to detect anomalous behavior [2]. edX relies on a network of partner universities that enforce institutional policies and conduct manual reviews of flagged submissions [2]. Udemy, prior to the merger, delegated integrity oversight to individual instructors, a model that the merged entity is now revising to include centralized detection tools [2].
Mechanisms of AI-Assisted Cheating
The cheating workflow typically begins with a learner providing a course prompt or assignment description to a generative-AI model. The model then produces a complete response—such as a research summary, programming solution, or analytical essay—within seconds [4]. In some cases, students employ “agent” scripts that automate the interaction with the AI, retrieve the output, and upload it directly to the learning management system [1].
Detection methods have evolved alongside the cheating techniques. Coursera’s AI-powered proctoring system analyzes visual and audio cues for signs of non-human interaction, while also cross-checking submitted text against known AI-generated patterns [2]. edX’s approach combines institutional policy enforcement with statistical analysis of response timing and similarity metrics, though partners have expressed concerns about false-positive rates [2]. Udemy’s post-merger strategy includes a hybrid model that leverages both automated scanning and instructor-led verification [2].
Immediate Impact on Learners and Institutions
AI Agents Complete Online Courses for Students, Prompting Concerns Over Degree Credibility
The prevalence of AI-generated coursework is prompting questions about the credibility of online credentials. Employers and accreditation bodies are reviewing the rigor of MOOC-based certificates as reports of automated cheating increase [1]. For students, the practice creates an uneven playing field, granting advantage to those who can access sophisticated AI agents while penalizing peers who adhere to traditional study methods [3].
Universities are responding by tightening enrollment verification, requiring multi-factor authentication for assessments, and integrating AI-detection software into their grading pipelines [4]. Online platforms are allocating additional resources to develop real-time monitoring tools and to train staff on interpreting detection alerts [2]. The combined effect is a rapid escalation in administrative workload and a shift toward assessment designs that emphasize oral exams, project-based work, and in-person verification to reduce reliance on text-based submissions [3].
In some cases, students employ “agent” scripts that automate the interaction with the AI, retrieve the output, and upload it directly to the learning management system [1].
What: AI agents are completing entire online courses for students, raising concerns about degree value.
When: The trend accelerated after generative AI tools became widely available in late 2022 and intensified following the Coursera-Udemy merger in May 2026.
Impact: Students, educators, and institutions face immediate challenges to maintain academic integrity and the perceived credibility of online credentials.
News
News
News
