In recent years, competition in AI development has intensified around the world. In particular, the growth of the United States and China is remarkable, and the EU (European Union) has established the “General Data Protection Regulation (GDPR)” as a countermeasure against privacy issues arising from ” big data ” that is indispensable for the growth of AI . .
As the development of artificial intelligence that is beneficial to society intensifies, the issue that comes up is the story of “after” when the AI era has permeated.
One of the issues that must be considered beyond the singularity of the era called “Singularity” is the issue of “human rights” in AI .
This time, we will introduce in detail the issues and initiatives related to AI x human rights.
What exactly are “human rights”?
According to the official website of the Ministry of Justice, human rights are described as follows.
History of human rights development
Let’s dig a little deeper into human rights.
The issue of human rights is still often taken up as a problem of racial discrimination. I will guide you through history.
Babylonian Human Rights Charter
Around 500 BC, the Cyrus Cylinder (Cyrus Cylinder Seal) left by Cyrus II, the first king of the Persian Achaemenid dynasty, is said to be the first declaration of human rights in history.
This cylinder of baked clay is recorded as proclaiming the emancipation of slaves, freedom of religious choice, and equal treatment of all races.
Magna Carta
Going back in time, the charter “Magna Carta” enacted in 1215 during the reign of King John of England (now England) is said to have had a great influence on the idea of human rights.
This charter is said to be the first in the world to limit the powers of the King of England, and to codify the protection of property and the right of the Church to be free from governmental interference, and is the basis of the current British Constitution. .
bill of rights
The next big influence was the Bill of Rights enacted by the British Parliament in 1689.
Based on the absolute premise of the existence of the King of England, it enacted rights that could only be enjoyed by a parliament and people who pledged loyalty to the King, and confirmed the rights that the people had inherited since ancient times.
French Declaration of Human Rights
The “French Declaration of Human Rights” was the result of the dissatisfaction of the bourgeoisie and the poor in France under the absolute monarchy.
A major feature of this Declaration is that the King did not put forward the Declaration of Human Rights. The French Revolution was triggered by Maximilien Robespierre on behalf of the people as dissatisfaction with the privileged class exploded.
The idea of human rights in artificial intelligence
Human rights issues in artificial intelligence can be categorized into two main categories.
- bias problem
- Ethics and liability issues in AI
I will explain each problem with a specific example.
The Bias Problem: The Filter Bubble Problem
The issue of bias is an essential issue when dealing with AI technology. When developing national or corporate services, imposing arbitrary biases on developers is a problem because it increases the risk that society will be constructed according to the convenience of developers.
And especially in the introduction of AI in a country-building where the government, not the citizens, is the main actor, it is necessary to attach importance to this bias issue.
What is particularly problematic in modern times is what is called a “filter bubble”. A “filter bubble” is a state in which news articles filtered by AI profiling make it seem as if the user is trapped in a small bubble.
It is believed that researchers are concerned that providing content that matches the user’s tastes and preferences will create more bias and that individuals will be trapped in a closed “world”.
Ethics and Responsibility Issues in AI: Autonomous Driving Trolley Problem
Do you know the “trolley problem” that is often cited as a problem related to human ethics?
When considering autonomous driving, it is necessary to consider cases such as the “trolley problem”. In other words, when requesting the AI itself to choose a life, it is necessary to decide what kind of judgment to make.
In this case, it is assumed that it would be extremely difficult to set a global standard on what criteria should be used for AI to select.
In addition, there is a good chance that medical mistakes will occur when AI is used in the medical field. Who will be held responsible then? The enactment of laws that prevent humans from blaming people is an issue that the future demands.
Issues and Efforts of Each Country Concerning AI Human Rights
I would like to summarize here the measures taken by countries around the world and the efforts of companies on how to face AI in the future. This time, we will introduce the initiatives in the following four regions.
- America
- China
- Europe
- Japan
I will explain each.
America
For the United States, which has the hegemony of AI ahead of others, it is already facing the problem of human rights. For example, there is the problem of misrecognition of gender in the face recognition system of AWS.
In the United States, AI is also being used in judgments. One example is a recidivism prediction program called “COMPAS”.
COMPAS is an AI that makes the defendant answer 137 questions and determines whether the defendant will reoffend or not by building a model with machine learning based on the results.
Since the recidivism rate system differs depending on race, “COMPAS” has become a problem because the developer clearly states biases such as race.
China
AI research in China is also active, and now it has become an AI advanced country along with the United States. One of the most notable examples of AI utilization in China is the credit score called “Sesame Credit”.
Credit score is used in various situations such as whether or not you can receive services depending on the score, credit card screening, and sharing services.
This credit score may be gradually introduced in Japan, but it can be said that there is also the risk of threatening human rights due to the opacity of the algorithm and privacy issues.
Europe
The EU (European Union) is most famous for the GDPR (General Data Protection Regulation). This is said to be a rule enacted by GAFA, a major American IT company, to prevent unauthorized use of personal information.
The GDPR is said to be one of the most advanced data protection regulations in the world, but it is said that the Jewish massacre in the Holocaust is famous as the background.
It is said that the Holocaust happened because the protection of personal information was not protected at all. Therefore, it can be said that the lessons of the Second World War strengthened the protection of individual rights.
Japan
In Japan, there is no particularly conspicuous use of AI, but the government has advocated the use of data and the introduction of AI technology, such as the “Principles of AI Development” and the “Osaka Track.”
Among them, the most noteworthy is the “Chiren society” that is dealt with in the “AI Development Principles”. The English translation is (Wisdom Network Society [WINS]).
This Chiren society is a society in which humans coexist with AI networks, freely and safely create, distribute, and connect data, information, and knowledge to form a “wisdom network”. It is a vision of a human-centered society in which cooperation beyond space between things develops and creative and vigorous development becomes possible.
In other words, the goal may be to treat AI as a thing and gradually increase its affinity.
at the end
Human rights issues centered on artificial intelligence are about to be considered as one’s own issues.
By looking back at history, it may be a good idea to think about how we will face AI in the distant future, and the issues of human rights and laws that may be obstacles.
