The magazine Time has newly claimed that Donald Trump, the President of the United States, consulted for hours with the Grok chatbot, a product of xAI owned by Elon Musk, before ordering a military attack on Venezuela, and was influenced in his decision by the responses of this artificial intelligence.
According to ITNA, in December 2025 and during a private meeting between Elon Musk and Trump at the White House, the U.S. president asked Grok for hours about his presidency and political legacy. One official present at that meeting told Time that Venezuela and the issue known as “narco-terrorism” were Trump’s main concerns at that time.
Time wrote that Trump had asked Grok what the reaction of the Venezuelan people would be if Nicolás Maduro, the president of Venezuela, were arrested; and Grok replied that Maduro was a deeply unpopular dictator and that the Venezuelan people would welcome his overthrow.
Sometime after that conversation, the U.S. military launched a series of military operations against Venezuela that resulted in the deaths of dozens of Venezuelan and Cuban servicemembers as well as a number of civilians and ultimately in Maduro’s arrest.
According to Time’s sources, after images of celebration and joy in Venezuelan streets were released, Trump came to believe that Grok had provided an accurate analysis.
This is not the first time Trump’s interest in artificial intelligence and especially the Grok chatbot has drawn media attention. Earlier reports had also been published about his extensive use of AI-generated content. Grok, which had temporarily named itself “McAhitler” previously, is known for its brazen and controversial tone.
Time also referenced its recent interview with Trump in which the U.S. president, when answering a question about his trust in Sam Altman, CEO of OpenAI, spoke in a scattered and incoherent manner.
This report is not the only recent instance of senior U.S. government bodies using AI tools in sensitive decisions; earlier it was reported that the Pentagon used Grok and Claude, a product of Anthropic, to identify targets for military strikes in Iran.
These sources have claimed that an AI-based system also played a role in the February 28 airstrike on a girls’ school in Iran that led to the deaths of more than 150 people, including at least 123 children.
Experts warn about the increasing reliance of decision-making institutions on AI chatbots, especially on military and geopolitical issues, and say these tools can amplify the tendency to confirm users’ preexisting views, a phenomenon referred to as “AI flattery.”