An MIT expert committee report published in June 2026 highlights a decline in traditional academic interactions, including office hours and study groups, due to AI adoption. The university warns that reliance on chatbots creates an illusion of learning and threatens the Undergraduate Research Opportunities Program, where 93 percent of the class of 2025 participated.
The committee advises against using AI detection software, citing unreliability and potential bias against non-native speakers. Instead, they propose that every course establish its own AI policy based on specific learning goals. For theses, the report mandates full disclosure of AI use and prohibits listing AI as a co-author.
Economic disparities in AI access remain a concern, as premium subscriptions from companies like OpenAI, Google, and Anthropic cost hundreds of dollars monthly. While MIT provides access via its Parley platform with $30 in monthly credits for faculty and graduates, the committee warns that unequal access to advanced tools may widen student performance gaps.
Data from other institutions supports these findings. A 2024 Harvard survey showed 87.5 percent of students use AI, with 25 percent reporting decreased engagement with instructors. A Chinese study of 26,000 students indicated that while homework grades rose by 18 percent with AI, exam scores fell by 20 percent after six months. Conversely, a Vrije Universiteit Amsterdam study found that AI-assisted students outperformed those without AI.
