In early September, an AI researcher at Anthropic resigned due to his belief that his leadership and colleagues were pursuing AI development in a dangerous and reckless manner. Following the growing number of whistleblowers sounding the alarm on what’s going on in the world of AI development, the public opinion on the risk of artificial intelligence has shifted in recent months in a largely bipartisan manner favoring stronger regulations for these companies rather than proceeding with the advancement of AI without stringent regulations. That said, the rhetoric coming from the White House doesn’t necessarily coincide with the American populace; in fact, it appears to contradict it. 

The White House continues to push the narrative of the AI Race, releasing a 90 point plan aimed at actions being taken to beat China in the race for AI supremacy. But unlike the Space Race which captivated American attention for over a decade and was a source of pride and motivation for so many, this rhetoric being pushed by Trump seems not to be landing as strongly for the average American. The disconnect between the demand from the public for regulation and the lack of action from the Trump Administration and politicians across the United States on who continue to fail to put meaningful legislation or regulatory oversight into the development of volatile AI technology (despite strong collective action and organization taking place nationwide) can be seen as an early example for the argument of American democracy eroding in the Trump Administration.

To examine the lack of regulation in AI as an example of possible democratic erosion, a fitting lens to view it is through Juan Linz’s and Alfred Stepan’s analysis regarding political legitimacy and efficacy. Linz and Stepan assert that through a government’s ability or will to solve the problems that their electorate cares about (efficacy) they accept that of the current government and its institutions as the “least evil” (legitimacy). The question of efficacy is most relevant in the case of AI regulation given that the average American far and wide views democracy as the legitimate government and therefore expects it to effectively address their issues. That said, Linz and Stepan state that if the public opinion ceases to support all political actors within a democratic regime it is likely to lead to an erosion of legitimacy. 

Not to say that because there isn’t sweeping legislation due to some polls the American democracy is failing, but the case of AI regulation is interesting in that there is a genuine disapproval among a large portion of the electorate yet actions aren’t being taken due to the current regime’s choice to choose the interests of others over the general concerned electorate. 

Perhaps the lack of action being taken is more fitting in the context of the following excerpt from Linz and Stepan:  “In addition to being responsive to the broad demands of the electorate and to party membership, democratic governments cannot ignore the demands of key well-organized interests whose withdrawal of confidence can be more decisive than the support of the electorate” (Linz and Stepan). 

It is no coincidence that in the midst of the AI explosion data centers (AI’s physical manifestation) have been popping up across the country and perhaps we are seeing one of the most blatant examples of the type of scenario Linz and Stepan explain above before our eyes being pushed too far in one direction. For example, despite mass mobilization and increasing disapproval of data center construction, the Trump Administration continues to unequivocally support the rapid construction of these data centers. It appears that in a manner characteristic of a society undergoing democratic erosion, the interests of well connected groups (such as corporations) are stronger than those of the electorate and the regime doesn’t necessarily care to hide this

Furthermore, largely in response to the plummeting public trust in the way AI development is being driven, the Trump Administration signed a “morally binding AI agreement”, with executives of the tech community. But given that the Trump Administration has given these corporations the green light to proceed with the full backing of the American government within the context of AI supremacy being a national security imperative, to what extent is the agreement simply a PR stunt? After all, no enforcement mechanism upon these companies is a dangerous precedent to set for AI as we have seen how corporations act when there is no oversight to enforce rules they will simply put profitability and the success of their business before all else. Fittingly, Linz and Stepan assert that “the success of a ruler might be based more on convincing society that the goals pursued are in its interest than in their actually being so” (Linz 19). It appears that the Trump Administration here is perhaps attempting to show an image of hedging to the whims of public opinion while still giving these organizations the green light to push on dangerously. 

We as Americans love being trailblazers. Westward Expansion, The Manhattan Project, The Space Race, and now it appears artificial intelligence. But we have to wonder if in pursuit of such a new groundbreaking endeavor, how much of our civic action is being undermined and ignored by those in power? How can one of the few possibly bipartisan policies we have in 2026 have be so blatantly disregarded by our politicians? Are we really willing to allow our collective mobilization, a cornerstone of American democratic culture, to be so obviously pushed aside by our leaders?