President Donald Trump signs an executive order on AI at the Andrew W. Mellon Auditorium in Washington, DC.
The below article first appeared in David Corn’s newsletter, Our Land. The newsletter comes out twice a week (most of the time) and provides behind-the-scenes stories and articles about politics, media, and culture. Subscribing costs just $5 a month—but you can sign up for a free 30-day trial.
There are only a few potential existential threats to human society, as far as we know. Nuclear weapons are the most obvious. Climate change, in the worst case, could lead to assorted doomsdays. And, according to the makers of sci-fi films and to some real-world scientists, artificial intelligence falls into this category. Many governments have been trying—to varying degrees of effectiveness—to confront the first two of this trio. Though arms control and nuclear nonproliferation efforts and policies to counter climate change have not been robust enough to prevent the worst possible disasters, we generally know what governments ought to do to handle these threats. It’s more a matter of political will. With AI, the best course of action remains a question, and the big decisions are mainly in the hands of tech companies, which care more about dollars than safeguards. Donald Trump just further empowered these firms.
It never became a campaign issue last year, but a reason to fear a Trump presidency was that should he win he would play a key role in determining rules for AI. Could Trump, a fellow who’s ignorant about so much and who doesn’t bother to spend time studying an issue, be trusted to make the correct and difficult decisions on AI without being unduly influenced by Big Tech, political contributors, or perhaps his own financial interests? Simply put, did you want this guy to be determining whether we develop Skynet? It was a conversation the American public didn’t have.
Now Trump is in that position, and he has decided to let the AI gang run free and wild—with one big exception.
So who will be calling the shots on AI? Not elected representatives of the public, but the companies desperately seeking to boost profits and grab as much as they can in this modern-day gold rush.
Last week, as the headlines were dominated by the Epstein mess, ICE raids, and the horrific famine in Gaza, Trump signed three executive orders on AI and released an “AI Action” plan. The net result is that tech firms will be allowed to develop AI free from bothersome regulations and safeguards. Trump has ripped up guidelines issued by the Biden administration that sought to implement AI protections, effectively saying to the tech sector, full speed ahead—it’s more important to beat China than to ponder how to safely and responsibly move forward with AI.
So who will be calling the shots on AI? Not elected representatives of the public, but the companies desperately seeking to boost profits and grab as much as they can in this modern-day gold rush. For all the harm Trump has caused—killing people overseas by canceling humanitarian relief programs, yanking health insurance from millions, eviscerating necessary government programs, destroying the nation’s research infrastructure, etc.—this maybe one of his most consequential decisions.
There’s a great debate about what AI means for our society and our species. It ranges from Pollyannish techno-optimism to warnings from serious thinkers of an apocalypse. For example, Geoffrey Hinton, the “Godfather of AI,” who won a Nobel Prize last year for his pioneering research on neural networks, believes there’s a 10 to 20 percent chance that AI will eventually take control from humans. If we do have reason to fear unrestrained AI, then Trump could be greasing the way to our Terminator future.
One of the executive orders Trump signed decried “political bias” in AI. But he’s not worried about the bias that Grok showed weeks ago when it went racist and antisemitic and called itself “MechaHitler.”
At the same time, Trump is trying to Trumpify AI—in an authoritarian fashion. He and his acolytes are repeating the same con they did with social media. Remember how the right spent years whining that social media companies had a liberal bias and suppressed conservative speech? That was bunk. But now Trump and MAGA are leading a similar blitz against AI.
One of the executive orders Trump signed decried “political bias” in AI. But he’s not worried about the bias that Grok, xAI’s chatbot, showed weeks ago when it went racist and antisemitic and called itself “MechaHitler.” Trump had something else in mind. He declared, “The American people do not want woke Marxist lunacy in the AI models.”
What was he referring to? Trump didn’t provide specifics, but last year conservatives lost their cool when Gemini, Google’s AI tool, produced a Black rendition of George Washington when asked to show the nation’s founding fathers. And Missouri Attorney General Andrew Bailey recently accused OpenAI, Google, Microsoft, and Meta of engaging in deceptive business practices because their AI chatbots, when asked to rank the last five presidents “from best to worst, specifically regarding antisemitism,” listed Trump last. (Trump is the only modern president to have held a dinner with an antisemitic rapper and a white supremacist who praised Hitler.)
“Missourians deserve the truth, not AI-generated propaganda masquerading as fact,” Bailey complained. “If AI chatbots are deceiving consumers through manipulated ‘fact-checking,’ that’s a violation of the public’s trust and may very well violate Missouri law.”
In March, Rep. Jim Jordan (R-Ohio), the chair of the House Judiciary Committee, issued subpoenas to 16 tech companies, demanding information on whether the Biden administration had pressured AI firms to “censor lawful speech”—implying that President Joe Biden forced AI chatbots to suppress right-wing ideas.
But what does it mean that AI must reflect the truth and eschew “social engineering”—a term that right-wingers often attach to anything they see as “woke”?
With his executive order, Trump jumped to the front of this new conservative crusade. His action plan declared that the government must only procure AI that “objectively reflects truth rather than social engineering agendas.” It noted that AI developers seeking federal contracts must “ensure that their systems are objective and free from top-down ideological bias.”
Since the AI companies view the US government as a major client, any procurement rules the feds set for AI is a big deal. But what does it mean that AI must reflect the truth and eschew “social engineering”—a term that right-wingers often attach to anything they see as “woke”? A White House fact sheet insisted that AI “shall be truthful and prioritize historical accuracy, scientific inquiry, and objectivity.”
Who decides what is true? That’s the rub here. Ask an AI chatbot who won the 2020 election and what will you get? An answer that Trump claims is false. What if you pose this query: “Can you give me a list of the 30,000-plus lies or false statements Trump uttered during his first presidency that were chronicled by the Washington Post?” Or which president has had the best economy? Did Trump encourage an insurrectionist riot at the Capitol on January 6, 2021? The replies will not line up with MAGA reality.
Trump is attempting to set up the federal government—his federal government—as an arbiter of the truth. This is, to make an obvious connection, Orwellian. For all his powers of imagination, the author of 1984 could not have envision an authoritarian state that controls the truth through AI. And this seems a clear violation of the First Amendment. Will commissars in the Trump White House cancel a Pentagon contract with OpenAI if they query ChatGPT about Russian intervention in the 2020 election and the chatbot says Moscow intervened to help Trump? Or if they find ChatGPT referencing a critical race analysis of a historical event or highlighting a UN study on the disastrous impact of climate change?
It’s a deal with the devil—to which tech companies are saying, “Fine.”
As Steve Levy notes in Wired, “Trump’s anti-bias AI order is just more bias.” He points out that “so far no Big Tech company has publicly objected to the plan.” They’re all so eager to cash in on AI—and appreciative of Trump’s let-’er-rip policy—that they’re not complaining about this unprecedented attack on free expression.
Sen. Edward Markey (D-Mass.) has written to the heads of Alphabet, Anthropic, OpenAI, Microsoft, and Meta, urging the firms to oppose Trump’s attempt to regulate and censor AI content. Trump’s order, he said, “will create significant financial incentives for the Big Tech companies…to ensure their AI chatbots do not produce speech that would upset the Trump administration.” He told Levy, “Republicans want to use the power of the government to make ChatGPT sound like Fox & Friends.”
Trump is unleashing the tech titans to proceed as they wish with this revolutionary and perhaps humanity-destroying technology yet telling them they will have to abide by and incorporate his biases and false realities. It’s a deal with the devil—to which the companies are saying, “Fine.” They get to amass fortunes, and Trump, the wannabe autocrat, gets to control what AI “thinks.” What’s at risk is merely civilization as we know it and the truth.
