The No-B.S. AI Law Blog

Does anyone else remember that viral Tom Cruise TikTok from 2021? The one where he did coin tricks and played golf? Many viewers knew it wasn’t real — it was a sophisticated deepfake that took months to create using high-end computing equipment, and even touted the username “deeptomcruise.”¹ Fast forward to January 2024: AI-generated robocalls mimicking President Biden’s voice targeted New Hampshire voters, demonstrating how synthetic media has evolved from entertaining social media content to a serious threat to our democratic process.²

The Rising Tide of Election Deepfakes

The New Hampshire incident wasn’t an isolated case. Research from the Brennan Center for Justice indicates that deepfake technology has become increasingly sophisticated and accessible, enabling the creation of convincing synthetic media across multiple formats.³ This includes:

• Voice clones that can mimic political figures
• Synthetic videos showing events that never occurred
• AI-generated images perfect for creating fake social media profiles
• Automated text that can flood platforms with disinformation

What makes this particularly concerning is that creating these deceptive materials no longer requires extensive technical expertise or expensive equipment. As noted in recent Reuters reporting, while high-quality deepfakes once demanded significant resources, the barrier to entry keeps getting lower.⁴

Beyond Individual Deception

The threat isn’t just about voters being fooled by single pieces of fake content. Recent studies have identified several distinct ways deepfakes undermine our democratic process:

1. Direct Voter Suppression: Like the NH robocall incident, deepfakes can spread false information about voting procedures or discourage participation.⁵
2. Last-Minute Disinformation: As seen in Slovakia’s recent election, synthetic media released just before an election leaves little time for fact-checking or correction.⁶
3. The Trust Crisis: Perhaps most insidiously, experts warn that the mere existence of convincing deepfakes creates what they call the “liar’s dividend” – where authentic evidence of misconduct can be dismissed as fake.⁷

The Social Media Amplifier

While we focus on obvious election interference, there’s a subtler threat in how deepfakes are being normalized through entertainment. Those viral Tom Cruise TikToks might seem harmless—and they’re clearly labeled as parody—but they’re making synthetic media feel routine, especially to younger audiences who may not always scrutinize content’s authenticity before sharing.⁸

Fighting Back: A Multi-Pronged Approach

The good news? We’re not helpless against this threat. Recent developments show promise on several fronts:

Legal Solutions

• The FCC recently ruled AI-generated voices in robocalls illegal⁹
• States like California are leading the way with laws requiring disclosure of synthetic campaign content¹⁰
• The bipartisan Deepfake Task Force Act aims to develop comprehensive federal responses¹¹

Technical Countermeasures

• Content provenance initiatives are developing ways to track digital content’s origin
• Social media platforms are implementing AI detection tools
• Research continues on “watermarking” synthetic content¹²

What Can We Do?

As we approach the 2024 elections, here are concrete steps we can take:

1. Support comprehensive regulation that addresses synthetic media threats while protecting legitimate uses
2. Demand platform accountability for detection and labeling systems
3. Share knowledge about deepfake detection and the importance of source verification

Looking Ahead

The 2024 election cycle will be a crucial testing ground for synthetic media’s impact on democracy. While the challenges are significant, our response to them will set important precedents for how we handle AI-enabled threats to electoral integrity in the future.

The solution lies not in fear but in proactive, informed action. By understanding the threats and supporting meaningful countermeasures, we can help ensure that AI serves to strengthen rather than undermine our democratic processes.

---

References

1. Tom [@deeptomcruise], “Sports!” TIKTOK (Feb. 22, 2021), https://www.tiktok.com/@deeptomcruise/video/6932166297996233989.
2. Huo Jingnan, It’s Quick and Easy to Clone Famous Politicians’ Voices, Despite Safeguards, NPR (May 31, 2024).
3. Daniel I. Weiner & Lawrence Norden, Regulating AI Deepfakes and Synthetic Media in the Political Arena, BRENNAN CTR. FOR JUST. (Dec. 5, 2023).
4. Mark Scott, Deepfakes, Distrust and Disinformation: Welcome to the AI Election, POLITICO (Apr. 16, 2024).
5. Todd C. Helmus, Artificial Intelligence, Deepfakes, and Disinformation: A Primer, RAND CORP. (July 2022).
6. Id. at 4.
7. Bobby Chesney & Danielle Citron, Deep Fakes: A Looming Challenge for Privacy, Democracy, and National Security, 107 CALIF. L. REV. 1753, 1785 (2019).
8. Weiner & Norden, supra note 3.
9. Huo Jingnan, supra note 2.
10. Gavin Newsom, Governor, Cal., Press Release, Governor Newsom Signs Bills to Combat Deepfake Election Content (Sept. 17, 2024).
11. U.S. Senate, Deepfake Task Force Act, S. 2559 (May 24, 2022).
12. Helmus, supra note 5, at 12-13.