Meta explains why its AI claimed Trump's assassination attempt didn't happen

Meta has addressed issues regarding its AI chatbot's responses to inquiries about an alleged assassination attempt on former President Trump. The company stated that the AI was programmed to refrain from commenting on recent events due to the potential for confusion and misinformation. However, in some instances, the AI incorrectly asserted that the assassination attempt did not occur, a phenomenon known as "hallucination," where AI generates inaccurate responses based on flawed training data. Meta acknowledged the need for updates to its AI's responses and is actively working to mitigate these hallucination issues.

Additionally, Meta clarified a mistake involving the application of a fact-check label to a photo of Trump taken after the alleged attempt. The label was mistakenly applied to the original image due to its similarity to a doctored version that misrepresented the scene. The company has since corrected this error.

Trump's supporters have criticized Meta for allegedly suppressing information related to the incident. In a related context, Google responded to claims made by Elon Musk regarding its search engine's autocomplete feature, which suggested "president donald duck" when searching for Trump. Google attributed this to a bug and emphasized that users can search freely. Overall, Meta continues to face scrutiny over its AI's handling of sensitive topics and the accuracy of its content moderation practices.