How To Assess AI-Aided Students?

Key Takeaways

• AI tools enable near‑perfect cheat sheets for take‑home exams
• Students exploit webcam tricks and smart glasses to bypass proctoring
• Oral midterms provide real‑time feedback and reduce AI misuse
• In‑person exams sidestep accommodation challenges tied to typing
• Law schools may shift final exams to oral formats permanently

Pulse Analysis

The rise of generative AI has upended traditional law school assessments, turning once‑secure proctored exams into vulnerable targets for sophisticated cheating. Students now employ physical hacks—such as vaseline‑coated webcams or discreet smart glasses—to capture exam prompts and feed them to tools like ChatGPT, receiving polished answers in seconds. This arms race forces educators to rethink evaluation methods, moving beyond static, written formats that AI can easily replicate.

One emerging solution is the brief, individualized oral midterm. By scheduling five‑minute one‑on‑one sessions, professors can verify a student’s analytical ability in real time, ask spontaneous follow‑up questions, and deliver immediate feedback. The format also simplifies accommodation logistics, as oral exams eliminate the need for typing or writing aids. Recording and transcribing each session adds a layer of accountability, ensuring disputes can be resolved with concrete evidence.

If oral assessments prove effective, they could reshape the broader landscape of legal education. Law schools might adopt in‑person oral finals, prioritizing demonstrable reasoning over written output. Such a shift would reinforce academic integrity while preserving the Socratic tradition that underpins legal training. Ultimately, the move signals a larger industry trend: institutions must adapt assessment design to stay ahead of AI’s disruptive capabilities.