OpenAI's ChatGPT Under Fire As Chinese Police Make First Arrest For AI-Assisted Fake News


27% profit every 20 days?

This is what Nic Chahine averages with his option buys. Not selling covered calls or spreads… BUYING options. Most traders don’t even have a winning percentage of 27% buying options. He has an 83% win rate. Here’s how he does it.


Chinese police have arrested a man for allegedly using OpenAI’s chatGPT to spread fake news about a train collision. 

What Happened: In what appears to be the first enforcement action under China’s unprecedented law concerning artificial intelligence, police in Gansu province have arrested a man for allegedly creating a fake news article about a train crash using chatGPT, reported CNBC.  

ENTER TO WIN $500 IN STOCK OR CRYPTO

Enter your email and you'll also get Benzinga's ultimate morning update AND a free $30 gift card and more!

According to the authorities, more than 20 accounts had posted the fabricated news story on a blogging platform owned by Baidu Inc. (NASDAQ:BIDU), which had amassed over 15,000 views. The accused reportedly used chatGPT to circumvent duplication checks and created versions of the article. 

See Also: Say Hi To ChatGPT’ Incognito Mode:’ Open AI Strives To Give Users More Data Control

It is pertinent to note that despite OpenAI’s AI language model being blocked by the country’s internet restrictions, a virtual private network or VPN can access chatGPT in China. 

The accused, whose surname is Hong, was arrested by the authorities under a newly implemented law regulating “deep synthesis technologies” introduced by China this year. The law explicitly prohibits the use of deep synthesis services, including AI-produced texts, images, videos or other media, to propagate fake news, the report noted. 

Why It’s Important: Earlier this year, 49 AI chatbot-generated news websites were uncovered by the news-rating group NewsGuard, raising concerns about the technology’s potential to facilitate fraudulent practices. 

Alex Keber’s investigation also uncovered a cluster of applications on the Mac App Store claiming to offer OpenAI’s chatGPT-like services. These apps are reportedly deceiving users into subscribing to costly services. 

Previously, cybersecurity experts warned that the latest version of OpenAI’s AI-powered model, GPT-4, may make it easier for bad actors to create new threats and build malware programs. 

Check out more of Benzinga’s Consumer Tech coverage by following this link.

Read Next: Google’s Latest Challenge To ChatGPT? Tech Giant To Reportedly Unveil Advanced AI Offering At I/O 2023


27% profit every 20 days?

This is what Nic Chahine averages with his option buys. Not selling covered calls or spreads… BUYING options. Most traders don’t even have a winning percentage of 27% buying options. He has an 83% win rate. Here’s how he does it.


Posted In: NewsTechMediaArtificial InteliigenceConsumer TechFake NewsOpenAi