When it comes to ethics, for some reason, relations between people immediately come to mind, and not, of course, the abduction of AI. However, in modern realities, ethical norms apply even to AI. From this article you will learn: What ethical principles exist in the development of AI and what may be the consequences of their absence. We give examples from real life that demonstrate how the violation of these principles can negatively affect society and certain groups of people.
Ukraine, which is a member of the Special Committee on Artificial Intelligence at the Council of Europe, joined the Recommendations of the Organization for Economic Cooperation and Development on Artificial Intelligence in October 2019, after which the Concept of the Development of Artificial Intelligence in Ukraine was approved by the Decree of the Cabinet of Ministers of Ukraine.
The latest changes in this legal field: in November 2023, Ukraine signed an international declaration on the safe development of AI at the AI Safety Summit in Great Britain.
This declaration, which was signed by another 29 countries, including the USA, Australia, EU countries, China and Saudi Arabia, provides for the development of AI in the interests of the state, business and citizens, with an emphasis on safety and ethical aspects of the use of this technology.
The main ethical principles in the development and implementation of AI are fairness, transparency and non-discrimination. Adherence to these principles is important not only from a moral point of view, but also to ensure public trust, legitimacy of technology and compliance with legal norms. In the future, I will supplement it with examples so that you better understand how the lack of these principles can negatively affect the work of AI in society.
Fairness in the context of AI means that systems must be designed to avoid bias and unfair outcomes. Transparency means that AI developers must ensure that it is possible to understand how their systems work.
Non-discrimination requires that AI does not harm specific groups of people based on characteristics such as race, gender, age, religion, etc. And now in more detail about why we should still adhere to ethical principles.
Ethical principles help protect basic human rights, ensuring fair and non-discriminatory treatment.
Next, transparent and fair AI systems increase the level of public trust in new technologies, which promotes their wider adoption. I will give one example of how the opacity of the work of AI can call into question the appropriateness of their use in general.
Amazon had an artificial intelligence that organized searchers’ resumes. He rejected resumes from women and instead offered to hire candidates who used masculine verbs in their language, even if these people did not have the required skills. This is sexism from AI.
In addition to this history, adherence to ethical principles helps to comply with the requirements of legislation governing the use of AI, reducing the risk of legal sanctions and fines.
Well, we are definitely talking about reputation here – companies that adhere to high ethical standards have a better reputation among consumers and partners, which can lead to an increase in their competitiveness on the market.
Yes, here, as with other norms and rules, chaos will ensue. The number of fake information will increase. Already, AI systems are capable of creating realistic texts, images, and videos that can be used to misinform, manipulate public opinion, and spread fake news.
As a result, trust in AI systems and the Internet as a source of information in general will decrease. As a result, the Internet will turn into an environment dominated by false information, which will make it less useful and uncomfortable to use.
The story dates back to 2016, when a scandal erupted over the COMPAS program, which is used by judges in some US states to assist in sentencing.
The program’s algorithms turned out to be racist. In her opinion, people with dark skin are more likely to reoffend than people with fair skin.
It is transparency that will help to find such built-in deviations and correct system distortions.
Whatever the algorithms are, they are likely to have built-in biases that are present in society in general and will be amplified. However, it is at least possible to work with these problems, observing ethical norms in the development of AI.
