The launch of Grok – artificial intelligence may be premature

X's chatbot is supposed to be an AI model that takes advantage of posts on the social media platform to display summaries of the latest news, but last week Grok's vulnerabilities were exposed for the umpteenth time, when the chatbot got confused and unfairly accused the NBA star of sabotage. .

“Klay Thompson has been charged in a series of brick attacks,” read the headline Grok created in a post that ran organically on X for several days. Under the headline, Grok went into detail to analyze fake news:

“In a strange development, NBA star Klay Thompson has been accused of vandalizing several homes with bricks in Sacramento. Authorities are investigating the case as several people claim that their homes were damaged, with windows smashed with bricks. Klay Thompson did not make a statement regarding the incidents and these incidents shook the local community, But no injuries were reported and the motive behind this unconfirmed vandalism remains unclear.

Grok appears to be confusing a common term in basketball where players are accused of unleashing a “brick” when they miss a shot completely. According to SF Gate, one of the first to spot Grok's mistake, Thompson had a rough night overall as he didn't miss any of his shots in his emotionally charged final game against the Golden State Warriors before he was left free to negotiate the next phase of his career.

There is also a disclaimer written in small print at the bottom of Grok's post, stating, “Grok is in its early stages and may make mistakes. Please verify the accuracy of his posts.”

But instead of confirming the accuracy of this particular post, some users chose to amplify the misinformation. Below the chatbot's post, statements from the alleged victims appeared. Some of these memes have registered millions of views.

In the past, both Microsoft and OpenAI have faced defamation lawsuits, with ChatGPT falsely accusing a politician and a radio producer of completely fictitious criminal acts. Microsoft has also been sued by an aviation professor who Bing Chat falsely described as a terrorist.

It's still unclear whether disclaimers like the one appearing under Grok would allow companies promoting these products to avoid legal consequences if someone decided to take action against them. Whether there is defamation will likely be determined by whether it can be shown that the platforms “knowingly” spread false information.

The fake news story with Thompson's alleged sabotage may have been the first time Grok's failure became widely known, but it wasn't the first time Grok spread nonsense by clicking on posts from X users who were joking on the platform. During the recent solar eclipse, for example, as Gizmodo reported, Grok published a headline that read, “Sun’s Strange Behavior: Experts Baffled.”

While it's fun to watch users mock Grok, the whole situation reveals that Grok may be vulnerable to manipulation by malicious actors to spread fake news or more dangerous propaganda.

The capabilities of AI models are still impressive, but there is growing concern about the quality of the data they are trained on, and the Grok and Klay Thompson “brick” incident adds to a string of similar situations where chatbots reproduce results completely incorrectly.