Lack of transparency – most AI tool developers (e.g., OpenAI, Meta, Google, etc.) will not provide detailed information on the dataset(s) used to train their AI.
Lack of verifiability – most AI tools will not provide references to the sources their output is based on, making it almost impossible to verify if those sources are accurate and reliable.
Lack of accuracy – data used to train AI models is not verified or reviewed for accuracy before it is used. There is no guarantee the data is correct; therefore, outputs can be completely incorrect.
Obsolete or outdated – training data is static, based on a point in time (in the past). Outputs can therefore be outdated or even obsolete depending on when you are using the AI tool.
