Dec 21, 2023 07:00:00

The number of websites posting AI-generated fake news has exploded by more than 1000% in half a year

High-performance AI such as ChatGPT and Bard can generate sentences that are indistinguishable from humans, but the content they generate is not necessarily true, and someone with malicious intent can generate fake news. can. It was reported that the number of websites posting such AI-generated fake news has rapidly increased by more than 1000% since May 2023, making it easier than ever for false rumors about politics, war, and natural disasters to spread. .

Tracking AI-enabled Misinformation: Over 600 'Unreliable AI-Generated News' Websites (and Counting), Plus the Top False Narratives Generated by Artificial Intelligence Tools - NewsGuard

https://www.newsguardtech.com/special-reports/ai-tracking-center/

How AI fake news is creating a 'misinformation superspreader' - The Washington Post
https://www.washingtonpost.com/technology/2023/12/17/ai-fake-news-misinformation/

According to a survey by NewsGuard, an organization that tracks online fake news, 614 unreliable AI-generated news sites have been confirmed as of December 2023. The number of websites posting AI-generated news has increased by more than 1,000% from 49 in May, and includes languages such as Arabic, Chinese, Czech, Dutch, English, French, German, Indonesian, and Italian. It seems that it exists across a total of 15 languages, including Japanese, Korean, Portuguese, Spanish, Tagalog, Thai, and Turkish.

These websites have names that make them seem like regular news sites, such as 'iBusiness Day,' 'Ireland Top News,' and 'Daily Time Update.' It is also unclear that the website operates with little human oversight or editing, and that the articles published are generated by AI.

'Some of these sites are generating hundreds, if not thousands, of articles a day,' said NewsGuard researcher Jack Brewster. He warned that AI could be used to spread fake news.

AI can generate articles on various topics such as politics, technology, entertainment, and travel, but the content is not based on facts.

An article discovered by NewsGuard stated that ``Israeli Prime Minister Benjamin Netanyahu 's psychiatrist had died, leaving behind a letter of protest implicating Prime Minister Netanyahu.'' thing.

Although the psychiatrist appearing in this fake news was a fictitious character, content based on this article was featured on an Iranian TV program, spread on media in Arabic, English, Indonesian, etc., and was also spread on TikTok. Fake news was also spread on SNS such as , Reddit, Instagram, and X (formerly Twitter).

Global Village Space, a website that publishes fake news about Prime Minister Netanyahu, covers a variety of political topics, including ``US sanctions against Russian arms suppliers,'' ``oil giant Saudi Aramco's investment in Pakistan,'' and ``US-China relations.'' We publish articles across serious topics. It also includes essays by experts at a Middle East think tank, a Harvard-educated lawyer, and website operator and Pakistani television newscaster Moeed Pirzada. Pirzada did not respond to inquiries from the Washington Post, a major daily newspaper.

Some of the articles posted on Global Village Space include essays and actual news, and it seems that fake news generated by AI is posted to blend in between them. As a result, it is becoming more difficult for readers to spot fake news. Jeffrey Blevins, a fake news expert and journalism professor at the University of Cincinnati, points out, ``Many people don't have the media literacy to tell that these articles are lies.''

In the 2020 US presidential election, a propaganda promotion group called Troll Farm based in Eastern Europe is said to have spread fake news and influenced the election results. It is also known that of the top 20 Facebook pages published for American Christians as of 2019, 19 pages

were operated by troll farms outside the United States .

However, while previous fake news was created manually, AI can generate huge amounts of fake news in an overwhelmingly short time and at low cost. Therefore, it is almost certain that websites like Global Village Space will proliferate during the 2024 presidential election, leading to a situation where AI-generated fake news will be disseminated in large quantities. At the time of writing, there has been no evidence that national intelligence agencies have used AI-generated news for propaganda, but Brewster said, ``I wouldn't be surprised at all if this was used. It will definitely be used.”

The Washington Post says, 'Historically, propaganda operations have relied on large numbers of low-wage workers and highly trained intelligence agents to build decent-looking websites. But AI Almost everyone, whether alone or a teenager in their basement, is producing content that is hard to distinguish from real news, making these websites easy to create.' .

It is difficult for governments to crack down on fake news content because it can violate freedom of speech, and the growing number of fake news websites remain largely unregulated. Blevins argues that the most important way to combat fake news is to 'increase the media literacy of the average reader.' ``Let's make people aware that there are websites like this in the world,'' he said, noting that they should not immediately trust a news site just because it looks serious at first glance, and that each news site has different credibility. It is necessary to understand.