By Ekaterina Koval.
Sc.D., Chief Investigator in the Middle-Volga Branch, The All-Russian State University of Justice in Sarans.
Big Data Ethics is a relatively new field of applied ethics, focused on understanding the practices that appear here and now in connection with the growth, accumulation and use of Big Data. There is no exact unit of measurement that allows to distinguish Big Data from the “small” one. The volume of Big Data is so large that it is impossible to process it with widely available software tools. Therefore, there are not many actors for which the issue of collecting, processing and storing Big Data is especially relevant. Usually these are governments and large corporations. But the problems of using Big Data concern almost all people.
So, we are faced with new risks and challenges. We are confronted with ethical issues at each stage of working with Big Data: collection, storage, processing and use of Big Data.
There are such problematic issues at the stage of collection Big Data. Who has the right to collect data? If Big User Data is being collected, how can informed consent be obtained from all? If data was collected for one purpose, and then another appeared, how to re-obtain informed consent?
The following assumptions can be made. Some moral norms impute certain qualities to a person. Thus, the categorical imperative imputes reasonableness and autonomy to a person. The golden rule of morality imputes trust to the Other. Perhaps a norm will soon appear that will impute an informed consent to a person to provide their data for the common good.
There are such problematic issues at the stage of storage Big Data. How long data can be stored? Can Big Data be transferred from one keeper to another?
If we admit that storage is a form of data existence, then, therefore, they are not only possible, but also necessary to be stored Big Data. In this case, there is a big privacy problem. In particular, there is a relationship between confidentiality and trust. The more we trust a person, the less we care about privacy by providing him with our data. But can we trust artificial intelligence, which alone can handle the processing of Big Data?
Next, consider the problems arising in the processing of Big Data. There are errors during data entry and errors during the development of data processing algorithms. For example, if a neural network designed for face recognition was taught mainly in photographs of white people, errors may occur in recognizing the faces of black people. If the algorithm is not designed correctly, then a «passionless» computer can discriminate against people. For example, an algorithm for finding employees for a company can discriminate against women. Thus, we are faced with a new challenge to justice. How to achieve justice in Big Data processing?
Finally, big problems arise on the stage of using Big Data. Today we can talk not only about the end of history or about the last person, but also about the end of anonymity. We are digitized to one degree or another. Therefore, our attitude to digitization is very important, as well as the attitude of those who collect, store, process and use our data.
As a rule, IT companies declare that they collect data for the benefit of the user or for the common good. But no one has yet canceled the rule that the main goal of a business is to maximize profit. Thus, modern society needs social norms that will not allow personal harm, protect their confidentiality. We need a new ethic, or perhaps just a rethinking of existing moral norms that will allow us to live in a Big Data world. This world will collapse without a man.
Artificial intelligence is much more powerful than humans, it has powerful analytical abilities. But we still cannot fully trust artificial intelligence. It does not have time to learn even on Big Data, because a person and society are changeable. For example, when a terrorist attack occurred in the metro of St. Petersburg, the Yandex taxi algorithm responded to increased demand by raising fares. The fare increased because people were trying to get out of the danger zone. And the algorithm could not respond to such a situation. As a result, the developers intervened and disabled the algorithm.
There are many other examples that make us think about the ethical issues associated with Big Data. For example, can hacked data be used? Should Big Data be open? What we must do in situations when the common good conflicts with the good of a particular user, and so on. Thus, we need a new ethic in order to live and develop in the Big Data and AI world.