What is Generative AI?

Generative Artificial Intelligence, often abbreviated as GenAI, generates text, images, multimedia, code, or other outputs using predictive models based on underlying patterns from large volumes of training data. A text-based GenAI platform, such as OpenAI’s ChatGPT, Google Gemini, Microsoft CoPilot, Anthropic Claude, DeepSeek, or Grok, is called a large language model, or LLM. GenAI became popular starting in November 2022, with the public release of ChatGPT, followed rapidly by the explosion in popularity of the GenAI image generators Midjourney and DALL-E.

Generative AI and Information Literacy

[This section was adapted from text generated by Google Gemini, a generative AI platform]

GenAI challenges our traditional understanding of authority and verification. Because LLMs are probabilistic, they prioritize plausibility over accuracy. When an AI does not know an answer, it may fill the gap with statistically likely—but factually incorrect—information. This phenomenon is known as hallucination. For example, an LLM might generate a perfect citation for a book that does not exist, attributing it to a real author, because that author often writes on that topic. We often do not know exactly what data a specific model was trained on, making it difficult to assess the provenance of the information it generates. In the age of AI, lateral reading (checking what other sources say about a topic) is critical. You must verify AI-generated claims against other sources of information.

Appropriate Uses of Generative AI in Coursework

The syllabus for your class should contain a policy on GenAI. If it doesn’t, you should ask your instructor for permission before using GenAI for your coursework. Your institution may have a recommended GenAI platform for students to use. In general, it is appropriate to use GenAI for the following purposes:

• Brainstorming: You can ask GenAI for help brainstorming topics for an assignment or brainstorming keywords to use in your research.
• Explanations: You can ask GenAI to help you understand a concept you are having trouble with or to summarize a passage from a reading that you are having trouble understanding.
• Outlining: You can ask GenAI to help you outline a paper or other assignment based on the topics you want to cover.
• Feedback: You can ask GenAI for feedback on your writing, such as places where it is unclear or needs more detail or better transitions. Keep in mind that writing you share with a GenAI platform may be used as training data for the platform.

There are also inappropriate uses of GenAI for coursework:

• Generating content: You should not turn in work generated by a GenAI platform, unless you have permission from your instructor and you have thoroughly reviewed and edited the content first.
• Misrepresentation: You should disclose your use of GenAI as a citation or note on your work; don’t present generated content as your own work, especially for assignments designed to evaluate your own skills.

The Many Controversies of GenAI

GenAI, like any relatively new technology, has generated many controversies:

1. Intellectual property: GenAI platforms are usually trained on publicly available data (such as websites). However, the authors and other creators who made that content publicly available usually did not provide permission for it to be used as training data. Therefore, many creators, especially those who rely on creating content as part of their livelihoods, such as authors and artists, consider the use of their work as training data to be theft and the generation of content using GenAI to be plagiarism. There’s also a question as to whether content generated by GenAI platforms is copyrightable. Currently the answer in the United States is “no,” but other countries are approaching copyright and AI differently.
2. Privacy: Any information that you submit to a GenAI platform may be used as training data. Don’t share confidential information and be thoughtful about sharing your own work, such as writing or art that you have created. GenAI is being built into ever more software platforms and web services (such as social media platforms), so read their privacy policies and consider opting out of having your data used as training data ⧉.
3. Environmental impact: Training GenAI platforms and building and maintaining the data centers required to run them requires massive amounts of electricity and water (for manufacturing and cooling). Before you choose to use GenAI for a task, consider whether it is really necessary.
4. Deceptive use: Not everyone who uses generative AI is using it in good faith or appropriately disclosing their use. Especially on social media, scrutinize comments, images, and videos for deceptive or misleading use of generative AI. Content related to political or controversial issues is a tempting target for bad actors who want to influence your opinion, and content creators may use generative AI to get your attention and/or your money.