Artificial Intelligence

It is important to recognize that there are limitations to using AI generated content. Here are some of the current issues to be aware of:

• GenAI systems are known to generate content that is false and inaccurate, referred to as hallucinations.
• GenAI systems may lack recent information and research in the training data.
• Content may be unevenly covered in the training data leading to unequal coverage.
• Not all GenAI systems provide detailed citations of the information provided.
• Training data, and therefore the system itself, may contain bias.

-Adapted from Fleming College Generative AI

A scholarly skepticism for the use of these applications is important for the following reasons:

• there are privacy and confidentiality concerns about personal user data;
• there is a potential for bias in generated output;
• there is a potential for misinformation, disinformation, and false information related to the generated output, including falsified references (referred to as hallucinations or generation of phantom data);
• there are copyright and plagiarism concerns related to the generated output;
• there is a potential risk for lack of diverse voices in the generated output;
• there are access issues related to superior applications behind paywalls;
• there is a potential detriment to the environment due to the energy (and harmful emissions) required to run these applications;
• there are potential concerns about its ethical use as it remains an unregulated industry.

-ICAI