Agentic AI for Teaching & Academic Work:
Judgement
Decide what should survive contact with the institution: multi-agent checks, grounding, guardrails, evaluation, responsible use, and adoption.
Chapters
Chapter 9Multi-agent systems
Use several perspectives to catch the errors one agent cannot see.
Chapter 10GroundingConstrain an agent to material you can inspect and trust.
Chapter 11Guardrails and the human gateDesign refusal, escalation, and human review into the workflow.
Chapter 12How would you know if it worked?Evaluate systems before they become institutional habits.
Chapter 13Responsible use in an academic settingConnect technical capability to policy and responsibility.
Chapter 14From workshop to practiceTurn a promising prototype into a useful academic practice.
This HTML reader gives you a searchable chapter guide. The complete typeset book remains available in the original PDF below.
Read Part 3
Use the embedded reader for continuous reading, or open the PDF in a new tab for printing and offline use.