When future historians look back and try to understand just how bizarre politics were in 2025, a useful period to revisit would be early July.
That’s when Elon Musk, shortly after his acrimonious split with the White House and accusations Trump was in the Epstein files, watched as his Grok AI anti-“woke” chatbot began calling itself “MechaHitler” and praising the genocidal leader.
While xAI quickly moved to take down the comments and ban hate speech posts, the offensive flare-up on a flagship product from one of the country’s richest and most politically connected figures captured the fast-moving, and often disorienting, ways politics and AI collided in 2025.
Top to bottom, AI impacted everything about national politics this year, from the president’s high-stakes trade negotiations with China, to the granular ways Americans communicated and sussed out fact versus fiction.
An (AI) picture is worth 1,000 words
No one in politics was more eager to embrace AI than Donald Trump, who built on early experiments during the 2024 campaign and proceeded to turn AI into his in-house “propaganda” factory, churning out countless, often outrageous, digital images to promote his policies and mock his enemies.
Accounts tied to the president, immigration agencies, and other top government figures now put out near daily AI-generated images and videos, though perhaps none were as striking as a September post, in which Trump made an Apocalypse Now-style image of himself as a general declaring “war” on Chicago, as the White House targeted the city for its next immigration crackdown.
“Chicago is about to find out why it’s called the Department of WAR,” the Truth Social post read, a striking message of aggression from a U.S. leader against an American city.
Mike Ananny, an associate professor of communications and journalism at the University of Southern California, said such images being shared at the highest level suggest a new political paradigm is underway. Being caught using doctored or fake images used to be considered shameful and verboten. Not so anymore.
“There’s no sense of, ‘Oh no, we were caught using a synthetically generated image,” he told The Independent. “All gloves are off. People don’t seem to care.”
Republicans weren’t the only ones hopping on the trend.
California Gov. Gavin Newsom, as part of his recent turn towards Trump-style online trolling tactics, now fires back at the MAGA crowd with near-daily AI images of his own, including a recent December post for “cuffing” season showing Trump, White House Deputy chief of Staff Stephen Miller, and Defense Secretary Pete Hegseth in handcuffs, set to a song from a 2022 Saturday Night Live sketch.
The same AI-fuelled dynamic played out in high-profile local races, including the harshly fought New York City mayoral race. There, former New York Gov. Andrew Cuomo’s campaign briefly released and then deleted an AI video featuring fake testimonials from “criminals for Zohran Mamdani,” his opponent, prompting criticisms that the former governor was using racist stereotypes. (The campaign later said the video was mistakenly posted by a “junior staffer.”)
Trump and the art of the AI deal
From the very first moments of the Trump administration, which began with major tech companies involved in AI lavishing donations on the president’s inauguration, artificial intelligence has been a core part of the White House agenda at home and abroad.
Domestically, AI drove huge parts of the Trump economic strategy.
To continue the boom in AI investment, the White House has pushed to waive landmark environmental laws like the National Environmental Policy Act to speed up data-center construction, while Trump, in December, signed an executive order seeking to block states from implementing their own patchwork of AI regulations. The administration has also touted the Stargate project, a project from OpenAI, Oracle, and SoftBank to direct hundreds of billions of dollars in domestic AI investment.
In Washington, Musk, at least before his break with Trump, was reportedly using AI to skim government contracts and cut regulations as part of his unprecedented, slash-and-burn DOGE effort to eliminate government spending.
Meanwhile, in the first three quarters of 2025, lobbying firms raked in nearly $100m working on AI-related issues, according to an estimate from Bloomberg Government, while millions more have been spent on in-house efforts from top Big Tech firms.
Even though Silicon Valley has leaned liberal in past decades, and the Trump movement had long criticized tech companies for alleged censorship, the simultaneous beginnings of the AI era and the Trump II presidency seemed to mark a political detente between Big Tech and MAGA.
Top AI firms are among the major donors to the president’s $400m White House ballroom project, and the president said in October a call from Jensen Huang, CEO of the major AI chipmaker Nvidia, was part of what convinced him not to send a massed force of agents into San Francisco for the kind of immigration crackdown that has besieged so many other Democrat-led cities across the country over the last year.
Overseas, major U.S. tech and AI companies had a way of being involved in key diplomatic moves, whether it was U.S. firms committing billions in investment and hardware as part of the September U.S.-UK “tech prosperity deal,” or the Trump administration using access to top U.S. AI chips as a bargaining tool, allowing sales of lesser Nvidia chips to adversaries like China so long as the U.S. got a 25 percent cut, while permitting top-of-the-line gear to go to friendlier countries like the UAE, where the Trump family has major business interests.
Fake calls, ‘glowfakes,’ and a shredding fabric of reality
Even as AI rapidly speeds up productivity in fields ranging from coding to medical discovery, the technology has quickly injected even further doubt into an already fractured political moment where Americans (including the president) regularly dismiss facts as fake news and believe demonstrably false information.
Beyond just the MechaHitler incident, Grok has appeared multiple times to align with the right-wing politics of its creator, despite Musk’s insistence that the AI chatbot is meant to be truth-seeking and outside political pressures.
In May, Grok began repeatedly invoking a non-existent “white genocide” in Musk’s native South Africa, telling users it was “instructed by my creators” to accept the genocide “as real and racially motivated.” (The company blamed the responses on someone making an “unauthorized modification” to Grok code.)
In June, Musk promised to modify Grok because it was “parroting legacy media” by pointing to accurate data showing right-wing political violence was more prevalent than left-wing attacks in recent decades.
AI has also blurred reality by giving tricksters and U.S. adversaries alike easy access to sophisticated impersonation tools.
In July, an impostor reportedly used a fake AI voice to impersonate Secretary of State Marco Rubio and contact top U.S. officials and foreign ministers.
Authorities believe the impostor was seeking to trick officials “with the goal of gaining access to information or accounts,” a diplomatic cable, obtained by The Washington Post, stated.
Daniel S. Schiff, co-director of Purdue University’s Governance and Responsible AI Lab, has worked with a team of researchers to catalogue a variety of different ways AI deepfakes are being used in political persuasion, from “fanfakes,” unrealistic positive AI tributes to political figures, to darkfakes, highly realistic and negative portrayals. An analysis from the 2024 election showed that fanfakes were, in fact, the most popular kind.
Schiff told The Independent there were countless ways AI could positively improve political information, such as making historical facts and political data easier to search and analyze, but the industry and public attention have been far more focused on rapidly using the latest tools.
“Hundreds of billions are being invested into innovation and a very tiny portion of that is being invested into safety and responsible approaches,” he said.
AI battle lines are being drawn for 2026
Given the deep skepticism many Americans feel towards AI, artificial intelligence looks set to be a key political issue in 2026, just like it was in 2025.
Silicon Valley firms have poured more than $100 million into a network of political action committees and groups to defend the industry amid midterm season, while political commentators argue that either party could benefit by answering Americans’ concerns about this new technology.
“One major question, going into 2026, is which party will speak for the Americans who abhor the incursions of A.I. into their lives and want to see its reach restricted. Another is whether widespread public hostility to this technology even matters given all the money behind it,” Michelle Goldberg wrote in a recent New York Times op-ed.
“We’ll soon start to find out not just how much A.I. is going to remake our democracy, but also to what degree we still have one.”
