As regulators struggle how to manage artificial intelligence without stifling innovation, the concept of an Algorithmic Bill of Rights, proposed by Professor Kartik Hosanagar, of the Wharton School of Business is receiving attention. Prof. Hosanagar, has been at the forefront of AI investigation years before it became a household name, publishing his book, A Human’s Guide To Machine Intelligence: How Algorithms Are Shaping Our Lives and How We Can Stay in Control, in 2019.
In his book, Prof. Hosanagar foresees so many of the issues that are front and center in 2023. Although Hosanagar covers a broad set of topics from a historical review of AI to the way AI influences almost every facet of our daily lives, his proposal for regulating AI is re-emerging as a topic of interest. Kartik has been quoted as saying “people tend to think of algorithms and machines as rational, objective decision-makers. In fact, we’re starting to see that the modern algorithms that are based on artificial intelligence tend to sometimes have biases and fairness issues.”
His concerns led to the concept of an AI Bill of rights. Hosanagar explains that he borrowed the AI Bill of Rights concept from the US Constitution which was drafted, at least in part, to protect citizens from a powerful government. The proposed AI Bill of Rights is a "self-governance contract between companies and citizens." Kartik argues that CEOs and corporations will be held publicly responsible for AI failures which will motivate the self-governance. Hosanagar's proposed AI Bill Rights includes the following highlights: Awareness that AI is making decisions that affect consumers or awareness that your data is used to train a language model. Transparency on how AI is being used in specific consumer transactions. Control, meaning having the ability for a consumer to take at least partial control over algorithms. Validations and quality assessment, such as including an inventory of machine learning models used and quality assessment to evaluate inputs and outputs.
Undoubtedly, as AI continues to dominate commercial and social interaction it is clear that regulatory and governance issues need to be addressed.