May 28, 2020
Ethical AI part 2: Human rights
Powerful tools need a solid base
AI can be a powerful tool. For example, in content moderation, AI can do much that’s good. It can remove hate speech before it gets published and ensure fruitful online interaction. At the same time, however, AI can be misused in many ways. Advanced, machine-learning-based AI learns to mimic each community’s publishing policy with high accuracy. If the policy is to decline all the critical voices, AI can be used irresponsibly in keeping them silent.
As a text analytics service provider, we consider it of utmost importance that the companies using Utopia AI have the privilege to control how they utilize it on their platforms. The Utopia AI is able to adapt to all languages, and it learns any unique moderation policy in just two weeks. It’s fair to say that Utopia AI is a very powerful product.
And a powerful tool needs a responsibly built, well-defined, solid base.
While the users of Utopia AI enjoy the freedom of defining their own moderation policy, we at Utopia Analytics focus on acting ethically. Since we understand that definitions of ethics are highly subjective, we have been trying our best to achieve mutual understanding not only internally but also with our customers. All things considered, we believe that the United Nation’s Universal Declaration of Human Rights (UDHR) is the most suitable foundation to safeguard the ethical use of Utopia AI.
The Declaration of Human Rights, which was historically adopted in 1948 by the United Nations General Assembly, consists of 30 articles affirming an individual’s rights, as elaborated in subsequent international treaties, economic transfers, regional human rights instruments, national constitutions, and other legal codes and standards.
The Declaration fits well with Utopia’s approach to content moderation, which is to focus on the protection of human rights online. By following these guidelines, we would be able to safeguard an extensive variety of rights, such as the right to information and the right to participation, without limiting it solely to freedom of speech.
Previous part: “Ethical AI part 1: Time to talk about responsibility“