“I created a slide deck for my Business Law 5 students to review the intersection of criminal and civil liability as part of a key Student Learning Objective. Rather than pulling from the broader universe of legal cases and concepts, I wanted the review to focus exclusively on the material we had already covered in class. To accomplish this, I used AI in combination with approximately 20 instructor-created resources that I uploaded into NotebookLM, including PowerPoints, legal cases, and supporting articles. I specifically instructed the AI to rely only on those uploaded materials when generating the presentation. The result was a concise 12-slide review deck that remained focused, aligned with course instruction, and less prone to AI hallucinations or inaccurate legal references. An added benefit is that the knowledge base now remains reusable. Because the materials are already uploaded, I can continue using the same NotebookLM workspace to quickly generate future review decks, study materials, and summaries, while also updating it with new cases and instructional content as needed.”
Kevin Gilligan Esq. Adjunct Professor, Business Law, El Camino College.
The Situation / Task
In Business Law 5, Professor Kevin Gilligan wanted to create a concise and engaging review resource to help students better understand one of the course’s Student Learning Outcomes: the intersection of criminal and civil liability in business law. Rather than building a new lecture review from scratch, he explored how AI could accelerate the process while keeping the material aligned with what students had already learned in class.
The Challenge
The primary challenge was maintaining accuracy and instructional alignment. Business law concepts can easily become overly broad, and publicly available AI tools sometimes introduce inaccurate information, fabricated legal examples, or “hallucinated” case references.
Professor Gilligan specifically wanted to:
- Limit the review content to the exact materials already covered in class
- Avoid introducing outside legal concepts or unsupported examples
- Reduce the risk of hallucinations and incorrect legal interpretations
- Save time preparing supplemental review content for students
How AI Helped
Professor Gilligan used Google’s NotebookLM to generate a customized review slide deck based entirely on his own instructional materials. He uploaded:
- Approximately 20 previously used lecture PowerPoints
- His course syllabus
- Selected articles and supplemental materials shared with students
He then prompted NotebookLM to create a presentation focused specifically on the intersection of criminal and civil liability in business law using only the uploaded content.
Within approximately one minute, NotebookLM generated a 12-slide review deck tailored to his course content and learning objectives. Because the AI was grounded exclusively in instructor-provided materials, the resulting presentation stayed closely aligned with classroom instruction and significantly reduced the likelihood of hallucinations or fabricated legal references.
Professor Gilligan noted that while some editing and validation were still necessary, the process reduced development time by an estimated one to two hours compared to building the presentation manually.
In a separate use case, he also used NotebookLM to generate a two-person podcast discussing Miranda v. Arizona. Students responded positively to the audio format and described it as a more engaging and accessible way to review foundational legal concepts. Like the slide deck, the podcast was generated solely from instructor-provided materials.
Example Prompt
“Create a slide deck reviewing the intersection of criminal and civil liability in Business Law based only on the uploaded materials. Focus on concepts and examples already covered in class and align the content to the course Student Learning Outcomes.”
AI Tool(s) Used
- NotebookLM
- Google Workspace for Education (El Camino College environment)
Outcome / Impact
- Generated a targeted 12-slide review deck in approximately one minute
- Saved an estimated 1–2 hours of preparation time
- Reduced hallucination risk by grounding AI outputs in instructor-provided materials
- Increased student engagement through AI-generated podcast discussions
- Created supplemental learning resources without expanding course prep workload
Lessons Learned / Advice for Colleagues
Professor Gilligan emphasized that the quality of AI-generated outputs depends heavily on the quality and completeness of the source materials uploaded.
Key takeaways:
- Uploading instructor-created materials improves accuracy and relevance
- Grounding AI outputs in course content helps minimize hallucinations
- Faculty should still review and validate all generated materials before sharing with students
- AI-generated visuals or explanations may occasionally oversimplify or inaccurately represent legal concepts
For example, Professor Gilligan noted that NotebookLM generated a “reasonable doubt” graphic that required correction because it was legally inaccurate. Even with those edits, the overall time savings and instructional value made the workflow worthwhile.
He also observed that El Camino College’s current free NotebookLM environment limits the number of slide decks that can be generated per day, and that expanded institutional access could further support instructional innovation.