Beyond Bans: Columbia Students Pioneer Critical AI Integration in Education

As artificial intelligence rapidly reshapes the landscape of education, a growing chorus of voices advocates for a shift in approach – moving beyond outright bans and embracing critical AI literacy. At Columbia University, a unique “AI Writing” course is leading the charge, challenging students to not only use AI tools but to dissect, critique and understand their implications.

The “AI-First” Classroom

Last fall, a discussion following a reading of a Recent Yorker article, “What Happens After A.I. Destroys College Writing?” revealed a surprising development on Columbia’s campus. Maximilian Milovidov, a freshman and member of TikTok’s Youth Council, shared that his “Writing AI” course stood out as the only one requiring, rather than prohibiting, the use of artificial intelligence.

This “AI-first” approach frames AI not as a threat to academic integrity, but as a tool to be understood and utilized responsibly. Milovidov draws a parallel to the introduction of the printing press, arguing that attempts to halt technological advancement are ultimately futile. He notes that today’s students will grow up in a world saturated with AI, with the majority of U.S. Teens already using AI chatbots for various purposes, including schoolwork.

Addressing Cognitive Offloading Concerns

A primary concern surrounding generative AI is the potential for “cognitive offloading” – the outsourcing of thinking to machines, which could weaken understanding. However, Milovidov’s experience in the “Writing AI” course suggests this risk is mitigated when students are actively encouraged to experiment with and critique large language models (LLMs). AI transforms into a dynamic study partner, offering both benefits and drawbacks.

The course structure involves students bringing their own ideas and drafts, feeding them into chatbots, and meticulously documenting the suggestions received. Crucially, students are then tasked with explaining why they accept or reject these suggestions. This process encourages critical thinking and helps students identify gaps in their own reasoning. Milovidov’s professor described this as the “friend test”: AI should provide feedback, not write the work itself.

AI as an Academic Leveler

Beyond enhancing critical thinking, AI tools can similarly help level the playing field in academia. For students lacking access to private tutoring, chatbots can provide practice questions, mock exams, and personalized feedback on writing. LLMs can also act as “study mode” tools, guiding students toward answers rather than simply providing them – mirroring the approach of a helpful teaching assistant. A 2025 Harvard University study indicated that students using an AI tutor achieved learning gains more than double those in traditional classrooms, and reported increased engagement.

The Importance of Critical Evaluation

Milovidov emphasizes that the inherent biases within AI systems – often trained on Western-centric data – underscore the need for students to question their outputs. He found that feeding drafts into an LLM often amplified existing weaknesses in his writing, highlighting the importance of human judgment and revision. The experience ultimately reinforced the idea that the ability to believe, judge, and revise is what truly distinguishes effective writers in an age where passable prose can be generated with ease.

Preparing for the Future of Work

The skills cultivated in the “Writing AI” course extend beyond the classroom. As AI increasingly automates entry-level jobs, the ability to collaborate effectively with these systems becomes paramount. Milovidov argues that education should prepare students to wield this technology, rather than simply endure its disruptive effects.

Maximilian Milovidov is a freshman at Columbia University studying Psychology & Information Science.