Artificial Intelligence (AI)

Unfortunately, AI models may be trained using biased information which can result in biased or discriminatory results. "Garbage in / Garbage Out" (GIGO) means that the output from AI is only as good as the input, or data it was trained on.

Even if the information used to train AI involves unconscious bias, this still affects the output. For more information, see:

There are a number of copyright issues relating to both the input and output stages of AI.

At the input stage, there’s the use of copyrighted materials to train LLMs. Does this violate copyright, or is it a transformative fair use (which means it would fall within the fair use exception to copyright law)? If the use of copyrighted materials to train AI does not violate the copyright law, are there still ethical/moral issues with AI using works for training purposes without the permission of the creators?

At the output stage, there are two issues:

1. because the AI model is trained on copyrighted material, if it generates output based on that material does the AI-generated material infringe on copyright; and
2. what, if any, copyright protection applies to works created by AI -- generally, a human creator is required for copyright to attach. What level of AI assistance is necessary before the work is created (for copyright purposes) by AI instead of a human? What rights do (and should) a creator have over works created with the use, or assistance of, AI?

AI also impacts the environment due to the energy and materials consumed, as well as the generation of electronic waste. For more information, see

AI may confidently assert information that is not true - such as when it creates and cites articles that do not exist. This is referred to as "hallucinations."

For more information, see:

AI can risk individual privacy in a number of ways. It may collect data without consent or permission, or be trained on datasets which contain private information which is loaded without concern for the privacy of the individuals the data refers to.

For more information, see: