ChatGPT Consumer in China Detained for Creating and Spreading Faux Information, Police Say
SAN FRANCISCO (AP) — Chinese language police mentioned they just lately detained a ChatGPT consumer for allegedly utilizing the AI-powered chatbot to create a pretend information story a couple of nonexistent practice crash. It is one of many first enforcement actions below a just lately enacted Chinese language legislation regulating AI-generated “deepfakes” — seemingly lifelike however fabricated digital photos, video or different media.
Based on a police report from the northwest Chinese language province of Gansu, a person recognized solely by his surname, Hong, used ChatGPT to create a pretend information article a couple of crash that supposedly led to the deaths of 9 development staff in Gansu. Twenty-one accounts on a preferred social platform, all owned by a media firm based mostly in southern China, unfold the pretend story inside a brief time period.
Like most overseas web sites and functions, ChatGPT is technically unavailable in China due to the nation’s “Nice Firewall,” which censors the web for residents. However decided people can acquire entry by way of generally out there “digital non-public community” software program that bypasses the firewall. The police report didn’t describe how Hong managed to make use of ChatGPT.
By the point Gansu safety officers realized the article was pretend, it had acquired 15,000 views, based on the report. Police subsequently raided Hong’s residence to gather proof after which took “felony coercive measures” towards Hong himself. Police use that phrase to explain non permanent measures to restrict the liberty of a suspect.
The brand new Chinese language deepfake legislation took impact on Jan. 10. It bans a number of classes of pretend media produced by “deep synthesis applied sciences” reminiscent of machine studying and digital actuality, however affords solely imprecise definitions for a lot of of those forbidden lessons.
Political Cartoons
Based on a translation of the legislation supplied by the crowdsourced website China Law Translate, it prohibits deepfakes utilized in actions that endanger nationwide safety, hurt the nation’s picture or societal public curiosity, or disturb “financial or social order.” It particularly prohibits using such applied sciences to provide, publish or transmit pretend information.
AP information assistant Caroline Chen in Beijing contributed to this report.
Copyright 2023 The Associated Press. All rights reserved. This materials is probably not printed, broadcast, rewritten or redistributed.