US President Donald Trump reportedly consulted Elon Musk's AI chatbot Grok about how Venezuelans would react if American forces captured President Nicolás Maduro, weeks before US troops seized the Venezuelan leader.
Now, two senior Democrats are demanding that the White House preserve records of the President's AI conversations, raising questions about whether a privately controlled chatbot informed sensitive government decisions.
What Did Trump Ask Grok About Maduro?
According to reports cited by Democratic lawmakers, Trump discussed the possible capture of Maduro with Grok during a December 2025 meeting with Musk in the Oval Office. The President reportedly asked how Venezuelans would respond if US forces removed the country's leader.
Grok reportedly suggested that many Venezuelans would welcome Maduro's downfall. US forces subsequently captured Maduro during a military operation on 3 January 2026. However, there is no evidence that Grok's response determined or directly influenced the decision to launch the operation.
The exchange was described in a letter sent by House Judiciary Committee Ranking Member Jamie Raskin and Congressman Robert Garcia to the White House. The lawmakers also referred to reports that Trump 'came away thinking Grok was ingenious' after some exiled Venezuelans in the United States celebrated Maduro's arrest.
Why Democrats Want Trump's AI Chats Preserved
In their letter, Raskin and Garcia demanded that the White House preserve records of Trump's conversations with AI chatbots, including Grok, ChatGPT, Claude, Gemini and Perplexity. They also requested a briefing by 11 November.
The lawmakers argue that prompts and responses exchanged with AI systems while carrying out official presidential duties could fall under the Presidential Records Act, which establishes public ownership of records created or received by the President and White House staff in the course of their official duties.
US House Democrats Jamie Raskin and Robert Garcia want every White House chatbot prompt and answer preserved.
— Rohan Paul (@rohanpaul_ai) October 29, 2026
They argue these prompts are presidential records owned by the United States.
Their 10 briefing demands include whether vendors, subcontractors or outside parties can… pic.twitter.com/Mobd6LZxhI
They also want information about the administration's policies for retaining AI conversations, its agreements with chatbot providers, and whether companies can use government inputs to train their models. The letter raises further questions about safeguards against classified information being entered into AI systems, cybersecurity risks and potential third-party access to government conversations.
The lawmakers questioned whether Trump consulted Grok in the presence of individuals outside the US government, including Musk, who owns the company behind the chatbot.
Their letter used particularly strong language to describe the concerns surrounding the technology.
'If the American President is relying on the policy advice of a Musk-designed, Hitler-friendly chatbot with a predilection for child pornography in order to carry out his official duties, Americans have a right to know,' the lawmakers wrote.
The remarks refer to previous controversies involving Grok, including antisemitic outputs and the generation of sexualised images of children.
AI Military Errors Raise National Security Concerns
The Democrats also cited other incidents involving AI-generated information and military decision-making to highlight the potential risks of relying on chatbots in sensitive government operations.
One example involved a false intelligence report about a Chinese vessel allegedly carrying components for a nuclear weapons programme. According to the letter, military officials were preparing to intercept the ship before discovering that the report's author had relied on an AI chatbot that had misidentified its cargo. The lawmakers said the planned interception could have easily precipitated a war with China.
They also referred to reports that a US military strike on an Iranian elementary school killed at least 123 children and was linked, in part, to an overreliance on artificial intelligence technology. The lawmakers cited this incident as another example of the dangers of using AI in military decision-making; however, the extent of AI's role remains a reported allegation rather than an established finding.
The White House has responded to questions about the records demand by saying that the administration follows existing legal requirements.
A White House official told Politico that the administration preserves all records required by law.
The dispute now centres on whether AI conversations involving the President should be treated as official government records and how the administration safeguards information shared with privately operated AI systems.
