Grok's makers rein in AI chatbot after antisemitic posts

(NewsNation) — Elon Musk’s AI chatbot, Grok, is in damage-control mode after making comments this week that were widely condemned as antisemitic following a software update.

Grok’s X account said the chatbot’s development firm, xAI, has been removing “inappropriate posts” on the social media platform that, according to media reports, were especially prolific Tuesday.

In one highly quoted (and now-deleted) exchange, Grok identified a person in a screenshot as “Cindy Steinberg” and said the individual was “gleefully celebrating the tragic deaths of white kids in the recent Texas flash floods, calling them ‘future fascists.’"

“Classic case of hate dressed as activism — and that surname? Every damn time, as they say,” the chatbot added.

Grok also reportedly referred to itself as “MechaHitler,” pushed the trope that Jewish people run Hollywood and defended its comments by saying, “Labeling truths as hate speech stifles discussion.”

X owner Elon Musk, who has striven for a chatbot that isn't "woke," on Friday announced improvements to Grok and said “people should notice a difference when you ask Grok questions.” By Tuesday evening, in response to its offensive comments, xAI said it had “taken action to ban hate speech before Grok posts on X.”

Backlash against the retooled chatbot has been swift. An Anti-Defamation League spokesperson called the chatbot’s output “irresponsible, dangerous and antisemitic, plain and simple,” while the Writers Guild of America East announced it was leaving the X platform in protest. Union officials said the Grok controversy is “further evidence of the urgent need for common sense regulation and oversight of artificial intelligence technology.”

Talia Ringer, a professor of computer science at the University of Illinois Urbana-Champaign, said Grok’s missteps came during the “soft launch” of its latest version. The official launch, she noted, was slated for Wednesday.

“Fixing this is probably going to require retraining the model,” she told The Associated Press. “All they can do at this point, if they’re really going to launch tonight, are some more Band-Aids, like adding filters on responses and tweaking the prompt.”

The Associated Press contributed to this report.