LibGuides: InfoSkills for Global Law: Generative AI: ChatGPT and other Large Language Models (LLMs)

Generative AI: ChatGPT and other Large Language Models (LLMs)

What is generative artificiaI intelligence (GenAI)?

Generative Artificial Intelligence, or GenAI, is a type of AI technology that can create original content – such as text, code, images, music, or videos – based on patterns it has learned from large datasets.

Many GenAI tools are powered by Large Language Models (LLMs). These are advanced AI systems trained on massive amounts of online data to understand and produce human-like language. Thanks to this training, LLMs can quickly generate coherent text and respond to prompts – short instructions or inputs that guide the model's output.

As a result, GenAI tools can answer questions, create written content, and support a wide range of language-based tasks with impressive fluency.

One of the most well-known GenAI tools is ChatGPT, developed by OpenAI. Other LLMs include:

CoPilot (Microsoft)
Gemini (Google)
Claude (Anthropic)
Grok (xAI)
DeepSeek (Baidu)

How students can use GenAI effectively

GenAI is a powerful tool with a wide range of applications, such as:

Providing useful suggestions for paper writing. AI chatbots like ChatGPT can assist you with:
- Brainstorming/discussing topic ideas
- Exploring different perspectives
- Refining your research focus
Recommending search terms for searching in Google Scholar or library databases
Providing summaries of lengthy texts and explanations of complex ideas
Providing transcriptions of spoken text. Several transcription tools exist that effectively convert spoken content from lectures and meetings into text, enabling review at a later time.
Translating text into different languages
Writing or debugging computer code

Challenges and limits of GenAI

Chatbots like ChatGPT are not designed to replace search engines. While they can help generate ideas and summarize concepts, they are not dependable for finding academic sources. They may produce fabricated or "hallucinated" citations – details about sources that don't actually exist. These citations often include convincing author names, article titles, journal names, and publication years, which can easily mislead users.

Example 1: Academic journal article

Smith, J. A., & Lee, R. T. (2019). Cognitive load and decision-making in high-stress environments. Journal of Applied Cognitive Psychology, 34(2), 145–162.

Issue: The journal name and article title sound plausible, but no such article or volume exists.

Example 2: Book

Thompson, L. M. (2017). The psychology of digital distraction. Harper Academic Press.

Issue: The publisher and topic seem credible, but the book doesn’t exist in any catalog or database.

Example 3: News article

Johnson, K. (2021, March 15). AI tools are changing the way we work. The Washington Herald.

Issue: The newspaper name is fabricated, and the article cannot be found in any archive.

Also, keep the following in mind:

Because large language models learn from historical data, they can sometimes reproduce biases or outdated information. Always review and verify AI-generated content carefully!
Copying AI-generated content directly is considered plagiarism. Tilburg University’s definition of plagiarism is:
- Using parts of a text written by someone else, or the reasoning or ideas of others for a thesis or other assignment, without due acknowledgement.
[The term acknowledgement refers to citing the original source.]
Plagiarism is wrong, dishonest, and can lead to serious consequences. So, think of GenAI-generated content as a guide, not the final product.
GenAI mainly learns from publicly accessible sources, so it doesn’t have access to deep web content like scholarly articles in library databases.
Use GenAI to support, not replace, critical thinking.
GenAI collects data. Make sure to protect your personal information!

Please note There are currently no university-wide rules on student use of GenAI. Always follow the guidelines provided by your lecturer.