Artificial Intelligence Bias and Fairness

2 min readSep 20


This image was created by Midjourney

AI is a recent and rapidly growing technology, and there is an explicit need for a code of ethics to be formulated and established for its ML (machine learning) models and systems.

These models work with biases. For example, when they associate a flower picture with its particular name. In this way, excluding bias from artificial intelligence is not feasible; instead, there’s a demand for bias development in the most suitable manner.

This development should be in accordance with principles based on fairness. Ethics in AI are as important as in any other aspect of human life. It’s a complex issue, but there are various solutions already working within these models.

Artificial intelligence systems can also be biased negatively; one example is when the data used to train these systems is not representative of the population that will use it. Another case is algorithmic bias, which occurs when the AI system is designed with an agenda. Finally, there’s also human bias, which happens when healthcare professionals or other users of these systems are influenced by their own interests. These can result in reduced accuracy, disparities in care, and loss of trust, negatively affecting the healthcare industry.

Some solutions that can address these critical challenges would require the design of fair algorithms, collecting and using representative data, developing oversight mechanisms, and training healthcare professionals. Some approaches will serve specific AI systems better than others, depending on their applications, so the best methods vary.

Nonetheless, by taking the measures above, we can assist in securing the healthcare use of artificial intelligence fairly and equitably. It’s a constant process, and as these systems become more complex, fairness and equity will continue to improve.

Ana Marques