The new rules state that algorithms should not be interfered with for the interest of an individual or a group in particular, and that companies need to guarantee transparency to the public to a degree that isn’t harmful to their business. (Image: Kakao)
SEOUL, Jan. 31 (Korea Bizwire) — Kakao has become the first South Korean AI developer to establish ethics guidelines concerning artificial intelligence technology.
The South Korean IT giant on Wednesday announced the release of the Kakao Algorithm Ethics Charter, which includes a set of rules concerning discrimination against social groups and ethical data collection and transparency regarding algorithms in AI technology development.
The charter sets out the goal of AI technology development projects as the ‘pursuit of happiness and in the interest of mankind’, and warns against the weaponization of AI data analysis against marginalized members of society.
The new move by Kakao comes on the heels of growing concerns over the threat of privacy invasion posed by AI technology, as the South Korean IT giant hopes to assure the public by promising to uphold social ethics when collecting, analyzing and using data.
The new rules state that algorithms should not be interfered with for the interest of an individual or a group in particular, and that companies need to guarantee transparency to the public to a degree that isn’t harmful to their business.
The charter has been published on Kakao’s official website, as well as in the January edition of online magazine Kakao AI report which was released on Wednesday, with a special commentary.
Kakao CEO Rim Ji-hoo says the charter represents the company’s principles and philosophy, and the will to develop algorithms in a transparent way.
“We will continue the efforts by each member to become an AI company that has a strong sense of ethics that society expects,” said Rim.
Calls have been growing within the IT industry for a similar set of ethics rules concerning artificial intelligence technology to be established.
Experts have said that unlike other types of technology, AI is hard to understand for the general public, and its self-learning feature can easily slip out of people’s control, and therefore tools need to be set up from the early stages of R&D projects to avoid both social and moral harm.
Internationally, Google’s announcement of plans to launch an AI ethics committee back in 2014 drew praise around the world.
M.H.Lee (mhlee@koreabizwire.com)
