Trump Says Zuckerberg Called Him
Let’s dive into a story that’s been making waves in the political and tech worlds.
Former President Donald Trump recently revealed in an interview with FOX Business' Maria Bartiromo that Meta CEO Mark Zuckerberg called him twice to apologize after Facebook mistakenly labeled a photo of Trump as misinformation.
The photo in question, which shows Trump raising his fist after surviving an assassination attempt during a campaign rally in Butler, Pennsylvania, was initially flagged by Facebook’s AI as misinformation. This labeling stirred up quite a bit of controversy, especially since the image went viral and was covered by news outlets around the world.
Trump shared with Bartiromo that Zuckerberg praised his response to the assassination attempt, calling it “brave” and acknowledging the mistake made by Facebook’s systems. According to Trump, Zuckerberg even mentioned that he wouldn’t be supporting a Democrat in the upcoming election out of respect for Trump’s actions—a statement that certainly raised some eyebrows.
Meta’s Vice President of Global Policy, Joel Kaplan, later issued a statement explaining what went wrong. According to Kaplan, Facebook’s AI mistakenly applied a fact-check label to the real photo after initially detecting a doctored version of the image.
The AI system, designed to flag content similar to what fact-checkers have already reviewed, inadvertently labeled the original photo due to subtle but significant differences between the doctored and authentic images. Kaplan emphasized that the mistake wasn’t due to bias but rather a technical error in the AI’s processing.
On top of that, Meta’s AI chatbot initially refused to answer questions about the assassination attempt, which further fueled concerns about the platform’s handling of such sensitive content.
Kaplan acknowledged the issue, explaining that the AI was programmed to give a generic response rather than risk providing inaccurate information. However, he admitted that Meta should have updated the AI’s responses sooner to address the situation more appropriately.
Adding to the mix, Google’s AI chatbot, Gemini, also refused to answer questions about the assassination attempt, redirecting users to Google Search for more accurate information. A Google spokesperson explained that this was in line with the company’s policy to restrict election-related responses on the Gemini platform, aiming to avoid the spread of misinformation.