Un cittadino cinese è stato arrestato dalla polizia per aver generato articoli falsi su un incidente ferroviario con ChatGPT, violando la legge.
Chinese authorities arrested a man for creating and publishing fake news with ChatGPT. This is the first arrest made under the new government guidelines that prohibit the use of artificial intelligence to spread fake news.
No false information with ChatGPT
The arrest was carried out by the police of Gansu Province (northwest of China). The man, whose last name is known only (Hong), used ChatGPT to generate fake articles about the railway accident, which occurred on April 25, which caused the death of nine people. Authorities found different versions of the same story (changing the location of the incident) on over 20 accounts of Baijiahao, a blogging service run by Baidu.
Hong is accused of using artificial intelligence technology to fabricate false information to be spread on social media. The man claimed to have used ChatGPT to rewrite a news item, bypass the Baijiahao filter that blocks duplicate posts and get a gain from Internet traffic. The fake items had exceeded 15,000 views prior to cancellation by the authorities.
China is one of the countries where access to ChatGPT is blocked, but citizens can bypass restrictions with a VPN. According to the Chinese police, Hong has committed a serious offense involving up to five years in prison. The guidelines introduced by the government also prohibit the use of technologies for the generation of images, audio and video in order to spread fake news. These contents must be clearly labelled to allow identification of the source.
Increase your SEO and SEM skills to position yourself on Google: discover the course for 19,90€.
Source: The VergeInizio modulo
Thank you, your comment is being approved
Thank you, your comment has been posted
Comment not sent
Thank you for subscribing to our newsletter
Oops, newsletter registration failed. Please try again.
Link copied to the clipboard
YOU MIGHT BE INTERESTEDFine modulo
China, government pushes for very stringent guidelines against AI
Create presentations in a flash with Decktopus AI
Midjourney does not work: the best alternatives
IBM Bets on AI: WatsonX and Other Projects Unveiled
Luca Colantuoni
Posted on 10 May 2023
Link copied to the clipboard