Deepfakes and Dark Money

In an age when even the very fabric of truth can easily be tampered with, AI-driven deepfakes represent a uniquely alarming threat to democracy. These doctored yet convincing videos and audio depict political figures uttering things they’ve never said, feeding public mistrust. Connected to dark money—money funneled secretly to influence elections—the AI fabrications serve as strong weapons; hence, any notion of fair and informed voting is put at unprecedented risk. This essay argues that the synergy between deepfakes and dark money threatens electoral integrity while actively dismantling the trust on which representative democracy is built.

AI advancements have made deepfake creation accessible and deceptively realistic. As mentioned by Richard W. Painter, all Artificial Intelligence needs is one image or just a small sample of someone’s voice to create a video of a political figure moving around, looking like them, and uttering just about anything they want. Unlike the pre-AI manipulation of media, today’s deepfakes require minimal sophistication and resources; thus, they have gone viral across social media platforms almost instantly. This kind of accessibility opens the floodgates for domestic and international actors alike to influence voter perception in ways that were previously unimaginable.

The Supreme Court’s ruling in Citizens United v. FEC drastically altered campaign finance by allowing nearly unlimited spending from political action committees, often funded by anonymous donors. Painter notes that this influx of “dark money” encourages various misinformation campaigns, including deepfakes, because there is no accountability regarding the origins of these funds. All that is needed is the financial backing provided by dark money, and these deepfakes will be spread abundantly and repeatedly across different media. They thereby take on legitimacy, swaying public opinion. In this case, anonymity allows bad actors—corporations, wealthy individuals, or even entities from other nations—to influence elections under a veil of secrecy.

The 2016 U.S. presidential election, whose results were widely reported to have been interfered with via Russian social media disinformation campaigns, set a particularly grim precedent for foreign interference in domestic politics. According to Painter, deepfakes give foreign powers even more effective tools for interference: fake messages and scenarios – indistinguishable from reality – can be created before an unsuspecting public. This implies that the actors, in combining dark money with AI deepfakes, actually disseminate disinformation and, far more importantly, introduce doubt into political discourse as being authentic. This result undermines the very idea of a free and fair election.

Deepfakes and dark money combine to destroy democracy in several ways. First, they undermine the judgment of voters who are subjected to false content presented to them through deepfakes as real. The inability of voters to separate fact from the falsehoods created by AI reduces them to a more vulnerable position—they are likely to be alienated by manipulated narratives and distanced from fact-based decisions. Painter adds that, for many years to come, public trust in both media and government will continue to erode, as individuals are bound to become less trusting of political messages, including legitimate ones. This ambiguity creates a vicious circle of distrust where voters are very inclined to give up voting and not take the outcome of elections seriously.

Furthermore, this damage extends well beyond single elections. The current stalemate in the FEC, combined with congressional inaction, means there are no immediate measures to curb these trends. As discussed by Painter, with no action, each election becomes a new playing ground for misinformation, and each cycle further enforces distrust.

Supporters argue that with continued improvements in detection tools, deepfakes can soon be overcome. However, as it is, technology remains far from being foolproof. The gap between improved detection tools and rapidly changing deepfake technology is too great—so are the limitations of social media sites to screen or label every piece of false content. Furthermore, once a person sees that video as realistic, it is going to be very hard for anyone—even independent fact-checkers—to guarantee full trust over any claims of falsification. Detection technology may help, but without more fundamental changes, it will fail to rebuild trust and inhibit the manipulative effects of dark money-driven misinformation campaigns.

By shattering the connection between public sources and political facts, AI-driven deepfakes and unregulated dark money have irreversibly changed the form of American democracy. By eroding the credibility of political information and undermining the reservoir of trust in public life, these forces are steering this nation toward a dark reality in which elections are determined not by informed citizenry or reason, but rather by who has raised the most money and equipped their investors with the best technology to count it. This threat needs to be immediately countered through action by Congress, the FEC, and tech platforms in the development of state-of-the-art detection systems and public awareness of AI-manufactured misinformation. Inaction is not an option; democratic principles demand proactive defense against a digitally engineered crisis of trust.

Sources

Painter, Richard W. Deepfake 2024: Will Citizens United and Artificial Intelligence Together Destroy Representative Democracy? Journal of National Security Law & Policy, 2023, page. 121–151. HeinOnline, 

(Chesney, Robert; and Citron, Danielle Keats. “Deep Fakes: A Looming Challenge for Privacy, Democracy, and National Security.” 2019, page. 882 – 892.)

Citizens United v. Federal Election Commission, 558 U.S. 310 (2010).