In April of last yr, america announced the ‘AI Ethics Guidelines’. In May 2019, the ‘Organization for Economic Cooperation and Development (OECD) Artificial Intelligence Recommendations’ got here out.

Increasingly more countries have been declaring artificial intelligence (AI) ethics since several years ago.

In December 2020, Korea also prepared ‘AI Ethics Standards’ on the 4th Industrial Revolution Committee directly under the President.

It is because AI technologies and services are inevitably inevitably colliding with humans, requiring some type of coordination.

The core content of the AI ​​ethical standards we prepared was the presentation of three basic principles and ten core requirements for human-centered artificial intelligence.

The three basic principles are the principle of human dignity, the principle of the general public good of society, and the principle of fit for purpose of technology. It’s transparency.

This yr, interest in ‘Chat GPT’ is hot. It isn’t any exaggeration to say that it’s the yr of ChatGPT. That is raising so many concerns.

There have also been cases where top trade secrets of domestic firms equivalent to Samsung and SK were easily exposed through ChatGPT. As well as, it was revealed that the Chinese e-commerce app ‘Pandoduduo’ stole users’ call logs, text messages, and photo albums.

These examples show that concerns that AI systems can collect and misuse user information can turn into a reality at any time. These were directly related to AI ethics.

AI ethics is an idea that focuses on social responsibility for the expected outcomes of AI models. It deals with ethical issues that will arise while interacting with society. A number of things to contemplate for this may be summarized as follows:

The primary is the issue of ‘prejudice’ in artificial intelligence. AI works based on data. If the information comprises bias, it will probably be reflected. This might cause major problems when AI takes over or influences human decisions. If left unchecked, it could expand and reproduce social inequality. This eventually results in the deterioration of the reliability of AI.

The second is ‘responsibility and transparency’. AI mainly serves to help human decisions and decisions. Nonetheless, AI can’t be held accountable. Responsibility for the outcomes of the choice should be borne by the developer or user. It is a problem that should be guaranteed to extend the reliability of AI.

Thirdly, there’s the problem of ‘privacy protection’. Many of the data handled by AI is inseparable from personal information. Examples include medical information, location information, preference, and usage information. With the intention to use such personal information, user consent should be obtained. Even when consent is obtained, it should be processed and used in order that user information isn’t exposed.

The fourth is ‘protection of human rights and safety’. All technology on the earth should eventually be for humans. This is strictly what the catchphrase, “AI for Humanity,” comprises. The identical goes for artificial intelligence. Technology that harms human rights or safety mustn’t exist on the earth.

In conclusion, AI ethics may be seen as a safeguard to seek out ways to coexist with humans while promoting the event of AI technology. Stuart Russell, who is known as the daddy of artificial intelligence, also emphasized that ‘real artificial intelligence is to create a system that coexists with humans’.

It’s for a similar reason that the OECD emphasizes ‘humanity’ as an AI development principle. The core theme of the EU’s AI regulation, which was originally scheduled to take effect in April, was also to thoroughly exclude harm to humans.

As well as, ISO/IEC JTC 1/SC 42 Artificial Intelligence Committee WG 3 Reliability Subcommittee, a representative standardization development organization, also published the concept of AI ethics. It’s ‘ISO/IEC TR 24368 Overview of Ethical and Social Concerns in Artificial Intelligence’.

The Korean government can also be preparing a typical template for collecting AI ethics cases as a national standard. We’re within the strategy of collecting opinions for this. It is a move ahead of the US.

Standard templates prepared in this manner may be developed into international standards in the long run. Through this, we hope that Korea will give you the option to act as a quick mover within the AI ​​field.

Cho Young-im, Gachon University Professor yicho@gachon.ac.kr