news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

IT Home reported on August 1 that according to a report by Nikkei today, Hitachi, Ltd. has developed technology that can determine whether an article was created by generative AI.

This technology is based on the articleWord usage patternsIn the future, it will not only be possible to judge whether a work is created by AI,Preventing the spread of misinformation created by generative AI, and can also help enterprises and administrative agencies write important documentsAvoid risks such as copyright infringement。

This newly developed technology will be embedded in the article creation technology based on the large language model of AI development companies, and the article creation mode will be defined as "rules"—— Use a specific word even if there are multiple synonymsIf the article uses a lot ofRule-based terms, it will judge that the article is created by AI. According to reports, Hitachi has also developedCombining multiple word selection rulestechnology, thereby improving the accuracy of judgment.

Currently, content generated by generative AI that is inconsistent with the facts (IT Home Note: commonly known as "AI hallucination") has also become a problem. If relevant technology can identify in advance that an article is created by generative AI, the risk of spreading false information can be reduced.