Please click here to opt in to receive via email our Roundup—brief daily summaries of news developments and commentary related to executive power.
Campaign finance regulation as conceived and pursued from the 1970s to the present day has mostly failed, and badly. It did not survive a range of pressures: one party’s implacable opposition, both parties’ complaints about the growing complexity of the statutory scheme, and changes in the Supreme Court’s constitutional jurisprudence that sharply limited congressional authority to restrict political money flows. Polarization played its part. As elections became close in margins and as political actors deemed their outcomes to be “existential” in significance, neither party—Democrats included—had much use for impediments to spending as much as they believed that they needed to win. And this has unleashed a free-for-all in spending, now amounting to billions of dollars, epitomized for many observers by the surge in “dark” (undisclosed) money to influence elections and the rise of the largely unregulated “super PACs.”
But the question of the role of money in politics—and certain kinds of money in politics—now assumes a different form than in the past and calls for carefully sculpted reforms to address a special case: the monies spent by the technology companies whose innovations and competition in the development of AI can transform the nation’s economy, its culture, and its politics. Companies like OpenAI, Google, and Anthropic have effectively limitless resources that they can channel in various ways, including through super PACs and unregulated “dark money” vehicles, to win elections. A number of them are now doing so, and this spending, and the companies and executives engaged in it, can be expected to steadily increase as the policy and political conflicts over AI intensify. Individual company executives have also contributed millions of dollars in support of their companies’ political programs. Forbes recently estimated that of the AI executives who gathered at the recent White House meeting to discuss industry self-regulation, ten had collectively donated $145 million in connection with the midterm elections.
Much of campaign finance law that might once have constrained corporate electioneering has eroded to the point of uselessness. But the companies in question are corporations of a particular kind, providing IT services to the government: they are federal contractors subject to special legal restrictions that, while astonishingly weak, could be substantially strengthened. The courts have left Congress more room in this than in other big-money spending contexts to fashion reforms that would address this type of spending. The prospect that the Supreme Court would uphold well-crafted federal contractor reform is reasonably good.
Since 1940, under an amendment to the Hatch Act, federal law has restricted political spending by federal contractors. A lead sponsor in the Senate declared that “the greatest source of corruption in American politics today is the use of money obtained from those who make profit out of contracts with the Government.” Congress’ action built on the original 1939 Hatch Act, which had constrained various partisan activities by federal government employees. In both cases, the overriding concern was the integrity and efficiency of the executive branch: its capacity to function competently and impartially, free of political influences and pressures.
The law as it stands today broadly prohibits any federal contractor from making—and any person from knowingly soliciting from a contractor—“any contribution of money or other things of value to any political party, committee, or candidate for public office or to any person for any political purpose.” The relevant regulations clarify that the prohibited spending includes “expenditures” as well as contributions. It is a “prohibition” largely in name only. Federal law generally prohibits corporations from contributing directly to candidates or parties, leaving this ban to affect primarily limited liability companies that have not elected corporate status for tax purposes, partnerships, and individual proprietors. But a contracting corporation is free to establish a “political action committee” or PAC with funds voluntarily contributed by its executives and administrative personnel, and use those monies to support candidates or parties. The Federal Election Commission has in the past construed the corporate restriction as narrowly as possible. In one such case, it concluded that the parent company of two LLC contractors was “a separate and distinct legal entity” able to contribute to candidates and parties without regard to the prohibition.
Moreover, AI executives may give personally, as may their spouses and family members; all employees may also make political contributions. They may also contribute in unlimited amounts to super PACs, and to support tax-exempt entities seeking to influence elections with “issue advertising” designed to influence voter choice.
These weak restrictions operate against a background fact: the corruption risks that moved Congress in 1940 have not receded. In fact, the problem is vastly more acute. There is striking evidence that the risks increase sharply when “political principals have a strong grip over their bureaucratic agents, in particular through appointees.” And “large donations to the party of the president substantially increase these risks, especially when the awarding agency is highly politicized (i.e., least insulated from the president).” With the advent of the unitary executive, and the experience now unfolding with the overlap of presidential control, weaponization and politicization, it would seem that the dangers of federal contractor politicking have risen to an entirely new level.
We are faced then with something of a perfect storm: a major policy commitment—the insulation of government contracting from political money—undermined by a law shot through with loopholes, the rise of the politicizing unitary executive, and a tech industry commanding trillions in the quest for political influence. The case for reform in this corner of campaign finance regulation is exceptionally powerful. And yet it may appear that the Supreme Court’s campaign finance jurisprudence, especially the affirmation of corporate spending rights in Citizens United, has closed the door on any meaningful reform initiative.
There is reason to believe otherwise. In Wagner v. Federal Election Commission, the United States Court of Appeals for the District of Columbia Circuit, ruling en banc, unanimously upheld the law against a challenge by three plaintiffs whose personal services contracts with the federal government subjected them to the prohibition. The plaintiffs alleged violations of the First Amendment and the guarantee of equal protection. In rejecting these claims, the court noted that even the plaintiffs had acknowledged that “Congress has greater latitude to restrict the expression of both employees and government contractors than it does with respect to the general public.” It emphasized in particular the government’s interest in protecting “against interference with merit-based public administration.” The court also indicated that the Congress might have significant flexibility in devising the means to that end, possibly “targeting only certain specific kinds of government contracting or doing so only during specific periods.” The Supreme Court declined without noted dissent to hear an appeal from the appellate body’s decision.
It is especially significant for purposes of the current reform issue that, in evaluating Congress’ regulatory interest, the court took account of the changing landscape it perceived in the field of government contracting. It noted that the “most relevant change…has been the enormous increase in the government’s reliance on contractors to do work previously performed by employees,” which meant that “if anything, that shift has only strengthened the original rationales” for the statutory ban. The government already spends billions of dollars on private technology services, and the share of AI within its total IT budget is soaring as agencies issued 2,255 identified AI-related awards between FY 2019 and 2025, with $4.1 billion in total obligations. This reliance on private industry will only continue to grow. President Trump refers to AI as “technology that will be bigger than the Industrial Revolution,” and those who run the leading tech companies as “the biggest people in the world.”
Reform legislation that replaces the largely useless prohibition now in place would certainly face energetic, well-funded resistance, from lobbying of the Congress to the constitutional challenge that would follow enactment. There is no doubt that the Supreme Court—this Court in particular—would closely scrutinize the reform’s nature and sweep.
Yet, in carving out more room for the regulation of this very specific kind of corporate political activity, the Justices who are committed to expansive constitutional protections for political speech and association would not be abandoning or seriously undermining their deregulatory agenda. A law that eliminates the current exceptions to the ban and expands transparency requirements would be a modest reform measure. It would apply to all federal contractors and should stand a fair chance of favorable constitutional review. The Congress could also clarify that the political spending ban applies to companies that provide uncompensated services to the government, as Meta does in supplying its “Llama” AI to federal government agencies.
It is even conceivable that, for the special category of federal contracting, Congress could go further and impose limits on certain other “independent spending” protected under Citizens United case law, such as executive and employee contributions to super PACs. The effect would not be insignificant: Elon Musk alone has donated millions to super PACs in this election cycle. At a minimum, the Congress could fashion special transparency requirements that would provide more visibility than at present into the AI interests behind much of this spending.
Legislative focus on one powerful sector of federal contracting might pass muster under an analysis like that of the Wagner Court, which pointed to the possibility that Congress might permissibly target “certain specific kinds of government contracting.” The Supreme Court may well decline to allow this targeted approach. Congress can only know if it builds the record for reform and makes its case.
Editor’s note: This piece is cross-posted at the NYU Law Democracy Project.
