Mark Zuckerberg Apologizes to Donald Trump for Facebook's "Mistake

Former President Trump revealed on "Mornings with Maria" that Meta CEO Mark Zuckerberg called him twice to apologize after Facebook initially labeled a photo of the former president raising a fist after an assassination attempt as misinformation.

Former President Donald Trump disclosed in an interview with Fox Business' Maria Bartiromo on "Mornings with Maria" that Meta CEO Mark Zuckerberg had contacted him to apologize after Facebook wrongly flagged a now-viral photograph of the former president.

The photograph, which depicted Trump raising a fist after a July 13 assassination attempt at his campaign rally in Butler, Pennsylvania, had initially been labeled as misinformation on Facebook.

Mark Zuckerberg Apologizes to Donald Trump for Facebook's "Mistake

"So, Mark Zuckerberg called me," Trump said. "First of all, he called me two times. He called me after the event, and he said that was really amazing. It was really brave. And he actually announced that he's not going to support a Democrat because he can't because he respected me for what I did that day. I think what I did to me was a normal response."

Trump went on to say that Zuckerberg "actually apologized" and acknowledged that Facebook had "made a mistake... and they're correcting the mistake."

Mark Zuckerberg Apologizes to Donald Trump for Facebook's "Mistake

In addition to apologizing, Trump claimed that Zuckerberg indicated that he would not be endorsing any candidate in the upcoming presidential election.

"He actually said they made a mistake... and they're correcting the mistake," Trump said. "And he said, 'By the way, you know, I'm not going to endorse anybody in the race. I'm not going to endorse anybody in the race because I don't want to affect the outcome, but I just wanted you to know that I respect what you did, and I think it was a great thing you did.' So, I appreciated that."

Mark Zuckerberg Apologizes to Donald Trump for Facebook's "Mistake

Meta Vice President of Global Policy Joel Kaplan later issued a statement explaining that Meta's fact-checking AI had incorrectly labeled the assassination attempt photo as misinformation due to its similarity to a doctored image that had been circulating online.

"Because the photo was altered, a fact check label was initially and correctly applied," Kaplan said. "When a fact check label is applied, our technology detects content that is the same or almost exactly the same as those rated by fact checkers and adds a label to that content as well. Given the similarities between the doctored photo and the original image, which are only subtly (although importantly) different, our systems incorrectly applied that fact check to the real photo, too. Our teams worked to quickly correct this mistake."

Mark Zuckerberg Apologizes to Donald Trump for Facebook's "Mistake

Kaplan also addressed concerns that Meta's AI chatbot had initially refused to answer questions about the assassination attempt, attributing it to the overwhelming amount of information during a breaking news situation.

"While neither error was the result of bias, it was unfortunate," Kaplan said. "Rather than have Meta AI give incorrect information about the attempted assassination, we programmed it to simply not answer questions about it after it happened... We've since updated the responses that Meta AI is providing about the assassination attempt, but we should have done this sooner."

Mark Zuckerberg Apologizes to Donald Trump for Facebook's "Mistake

Google's AI chatbot, Gemini, also faced criticism for refusing to answer questions about the shooting. Google later explained that this was due to restrictions on election-related queries on its Gemini platform.

"As we announced last year, we restrict responses for election-related queries on the Gemini app and web experience," a Google spokesperson said. "By clicking the blue link in the response, you'll be directed to the accurate and up-to-date Search results."

The incidents have raised concerns about the reliability of AI in moderating online content and the potential for bias in these systems.