Students are employing generative‑AI services that autonomously finish coursework and exams, while universities report rising detection challenges.
An investigation by the New York Times on August 10, 2026 documented that organized AI agents are being hired to enroll in, complete, and submit assignments for entire online courses on behalf of students [1]. The Business Standard reported a similar phenomenon in Bangladesh on August 11, 2026, confirming that the practice is observed across multiple continents [2].
The reports identify college students as primary users of these services, while faculty members, institutional integrity offices, and AI‑ethics researchers are responding to the emerging threat [1][3]. The rapid adoption is linked to the widespread availability of generative‑AI tools that can produce human‑like text, solve quantitative problems, and simulate interactive exam performance [4].
Scale and Operation of AI Cheating Services
The New York Times article describes a market where AI agents are offered through online platforms that guarantee “completion of the course with a passing grade” for a fee ranging from $200 to $2,000 per semester [1]. These agents employ large language models (LLMs) combined with custom scripts to navigate learning‑management systems, submit assignments, and even participate in timed quizzes by mimicking mouse movements and keystrokes [1].
The Business Standard corroborates the model, noting that in South Asian universities, students have accessed similar services via messaging apps and local “tech‑tutors” who act as intermediaries between the AI backend and the student’s account [2]. In both regions, the services claim anonymity, with payments processed through cryptocurrency to evade institutional monitoring [2].
These agents employ large language models (LLMs) combined with custom scripts to navigate learning‑management systems, submit assignments, and even participate in timed quizzes by mimicking mouse movements and keystrokes [1].
A faculty member who authored a commentary in Nature reported personal detection of AI‑generated responses during a welfare‑economics exam, citing unusually consistent phrasing and statistical analysis that matched known LLM output patterns [3]. The professor’s account illustrates how AI agents can be integrated into live exam environments, not just pre‑written assignments [3].
Feedough’s 2026 statistical review indicates that reported incidents of AI‑facilitated academic misconduct rose by 68 percent between January and April 2026, with online‑only programs accounting for the majority of cases [4]. The same source notes that traditional plagiarism detectors failed to flag AI‑generated text, prompting institutions to adopt new AI‑detection tools that still lag behind evolving models [4].
For students who do not employ AI agents, the reports suggest a competitive disadvantage, as peers using the services achieve higher grades with reduced effort [1][2]. Surveys conducted by university integrity offices in the United States and Bangladesh indicate that 22 percent of respondents suspect classmates of AI‑assisted cheating, influencing perceptions of fairness in grading [1][2].
Institutions are responding by revising assessment designs. The New York Times notes that several U.S. universities have shifted high‑stakes exams to in‑person proctoring, introduced oral defense components, and integrated AI‑output detection software into their grading pipelines [1]. In Bangladesh, the Business Standard reports that at least three major private universities announced mandatory on‑campus assessments for core courses and launched faculty training on AI‑detection techniques [2].
The broader implication for the credibility of online degrees is highlighted by the articles: accreditation bodies are reviewing compliance standards to include AI‑integrity protocols, and employers are being advised to verify candidate credentials through supplemental skill assessments [1][3].
Response Strategies and Future Monitoring Universities are allocating resources to develop multimodal detection systems that analyze writing style, timing metadata, and interaction patterns within learning‑management platforms [4].
Response Strategies and Future Monitoring
Universities are allocating resources to develop multimodal detection systems that analyze writing style, timing metadata, and interaction patterns within learning‑management platforms [4]. The Nature commentary emphasizes the importance of faculty awareness, recommending that instructors design assignments that require personalized data, iterative drafts, and in‑class presentations to reduce reliance on static AI outputs [3].
Policy makers in higher education are also convening working groups to draft guidelines for permissible AI use, distinguishing between assistive tools for research and prohibited services that complete coursework autonomously [1]. Early adoption of these guidelines may mitigate the erosion of trust in online credentials, according to the New York Times analysis [1].
Key Facts
What: AI agents are being hired to complete entire online courses and exams for students.
When: Reports published on August 10‑11, 2026, with rising incidents documented throughout 2026.
WestJet Airlines has agreed to pay C$4.5 million to settle a class action lawsuit filed by female flight attendants alleging sexual harassment and unsafe workplace…
NewsNews
