Two Democratic lawmakers demand to know if Trump uses AI to make decisions

Two Democratic members of Congress have called on U.S. President Donald Trump to retain records of any official correspondence with artificial intelligence chatbots, raising questions about whether Elon Musk’s Grok program could influence sensitive military decisions.

In a letter dated Wednesday obtained by Politico, Congressmen Jamie Raskin and Robert Garcia stated that questions posed by Trump and his aides in the White House, along with the chatbots’ responses, constitute presidential records that must be preserved under federal law.

They requested a closed-door briefing by November 11 addressing the President’s use of artificial intelligence, security safeguards, and whether third-party companies have access to sensitive conversations.

They wrote, “If the U.S. President relies on political advice from a chatbot designed by Musk, who is sympathetic to Hitler and has a propensity for child pornography, to perform his official duties, then Americans have the right to know that.”

This wording alludes to previous controversies surrounding Grok’s issuance of antisemitic statements and its generation of sexually explicit images of children.

The letter follows reports that Trump spent hours using the Grok program during a meeting with Elon Musk in the Oval Office in December 2025, including asking about the reaction of Venezuelans if U.S. forces arrested President Nicolás Maduro.

Reports indicated that Grok stated many would welcome Maduro’s downfall. Maduro was arrested by U.S. forces weeks later, but no evidence emerged suggesting the chatbot influenced the operation.

The lawmakers raised concerns about Trump’s alleged reliance on a privately operated AI chatbot for foreign policy matters, warning that AI-generated answers could be inaccurate or manipulated.

They cited reports of an intelligence error related to AI concerning a Chinese ship, and others alleging that overreliance on AI contributed to a U.S. airstrike on a school in Iran that killed more than 100 children.

The lawmakers added, “If the U.S. military kills dozens of innocent children based on incorrect advice from an AI-powered chatbot, we also have the right to know that.”

Neither of these incidents has been proven to involve Grok in those military decisions.

They also requested details on restrictions regarding the input of classified information, procedures for handling chatbot records, and any potential access by third parties.

Their request included other chatbots, such as ChatGPT from OpenAI, Claude from Anthropic, and Gemini from Google.

A White House official told Agence France-Presse that the administration retains all records required by law.