AI Endgame: Political Deepfakes

Newsletter #4

Oct 11, 2024

October 11, 2024

By Debbie Coffey, AI Endgame

Well, since the last newsletter, I’ve returned from my camping trip. The big “news” while I was away: as I entered a dark RV and stepped over one of my dogs (who laid down as soon as she entered), my foot almost landed on top of hers, so to avoid crushing her foot, I threw myself forward in a twisting motion and inadvertently pulled a gluteus maximus muscle. Since then, it’s been painful to sit and I’m walking around like I’m on stilts (but it’s all about the dogs).

Don’t tread on me.

Moving on to more important news, foreign countries are producing AI generated content to influence our 2024 Presidential election.

Russia, China and Iran are creating and distributing false information in order to shift U.S. public opinion and manipulate the decisions of U.S. voters to elect certain “desired” candidates (candidates that these countries believe can be swayed to benefit their own agendas). [1]

This month, U.S. intelligence officials warned that covert foreign influence campaigns will continue to question the validity of election results, even after our polls close, in an attempt to undermine our trust in our election’s integrity. Disinformation promoting the idea that there are U.S. voting irregularities is already being spread by Russian state media and Kremlin-friendly organizations. [2]

(In my opinion, these countries want to cause division among us and to undermine the strength of the United States and democracy.) Much of this is being done specifically by using deepfakes.

DEEPFAKES

Deepfakes are extremely realistic fake images, videos or audio recordings created by means of artificial intelligence. They are used to spread disinformation and to shape or influence public opinion.

“Using artificial intelligence, deepfakes can mimic a person's voice and facial features. The technology uses an audio recording of someone's voice to make it say things that the person might never have said.

It can mimic someone's facial movements from videos of them, or even just a picture of their face.” [3]

This technology is improving so rapidly that it’s becoming increasingly difficult for us to discern false images from real images. In earlier iterations, deepfakes revealed flaws like oddly shaped hands with extra fingers, or the edges of faces were blurry. Now deepfakes are creating “average” faces that don’t have any unusual facial features that appear on real people. For instance, the skin on faces used in deepfakes may be smooth and lack any wrinkles. Since people view average faces as being more trustworthy, people may more easily fall for the messages of the deepfakes. [4]

RUSSIAN POLITICAL DEEPFAKES INFLUENCING THE U.S. ELECTION

Imagine you’re browsing the internet one evening, and you see a video of a woman who claims to be the victim of a hit and run accident by Kamala Harris in 2011. You take what this woman says at face value, and you then wonder how we can trust a politician who flouts the law.

This video was described last month by an official from the Office of the Director of National Intelligence (ODNI), who said Russian influence actors staged a video in which a woman claimed to be a victim of a hit-and-run by Harris in 2011. There was no evidence this ever happened, and follow up investigations proved this did not happen. In fact, Microsoft learned that Russia was behind this video, which was spread by a website claiming to be a nonexistent local San Francisco TV station. [5]

There are thousands of AI generated deepfakes all over the internet, and we may never find out who created them.

Maybe you saw one of the hundreds of Tenet Media videos about immigrants or anti-white prejudice. Maybe you were outraged when you saw some of the Tenet Media content that Elon Musk shared on 60 different occasions with his 198 million followers.[6] If someone as smart as Elon Musk shares something, it must be true, right?

U.S. Prosecutors recently revealed that two Russian nationals working for RT (a media outlet formerly known as Russia Today, that is funded and controlled by the Russian government), funneled $10 million through a series of shell entities to a Tennessee-based company launched in 2023 (the company was unnamed in the indictment, but the description of this company matches Tenet Media) that publishes videos on TikTok, Instagram, YouTube and X (formerly known as Twitter).

Tenet Media posted about 2,000 videos on YouTube that had over 16.5 million views.

Tenet’s website listed six conservative social media personalities on its talent roster - Tim Pool, Benny Johnson, Dave Rubin, Lauren Southern, Taylor Hansen and Matt Christiansen. [7] These people unknowingly distributed Russian propaganda (to over 20 million of their followers and subscribers).

It sure seems like foreign governments are using AI for PsyOps on us, doesn’t it? (PsyOps are Psychological operations that use select information in ways that will influence people’s reasoning and beliefs, with the ultimate goal of influencing governments, organizations, groups, and foreign powers.) The “ways that will influence your reasoning and beliefs” include psychological profiling (focusing on your fears) and plans to create social divide and chaos.

Luckily, I’m not the only person who is fretting about this. The U.S. Government Accountability Office (GAO) issued a report to Congress “FOREIGN DISINFORMATION: Defining and Detecting Threats” (Sept. 2024), that explained foreign disinformation is spread in a variety of ways, including propaganda, social media, and artificial intelligence such as deepfakes. The GAO’s stance is that foreign disinformation threatens U.S. national security. [8]

TAYLOR SWIFT

To give you an example of the extent of the far reach of deepfakes (although the following example is not political), Taylor Swift was the victim of a deepfake in January 2024, when AI was used to generate FAKE, sexually explicit images of her that spread like wildfire over social media.

“One image shared by a user on X was viewed 47 million times before the account was suspended.” [9]

Viewed 47 million times? This number is more than the population of over 200 countries in the world, and this number also far exceeds the entire population of Canada. [10]

Widespread deepfakes are increasingly dangerous if you consider the quote “If you repeat a lie often enough, people will believe it, and you will even come to believe it yourself.” This quote was by Joseph Goebbels, Hiter’s minister of propaganda for the Nazi Party. [11]

Not all political deepfakes are by foreign influences. There was recently a deepfake on Donald Trump’s Truth Social account involving Taylor Swift, where AI-manipulated photos depicted Swift as Uncle Sam with the text, “Taylor wants you to vote for Donald Trump” and Swift fans wearing “Swifties for Trump” T-shirts. [12]

Taylor Swift responded to this deepfake by stating “Recently I was made aware that AI of ‘me’ falsely endorsing Donald Trump’s presidential run was posted to his site. It really conjured up my fears around AI, and the dangers of spreading misinformation…” [13] Swift then endorsed Vice President Harris.

And Taylor Swift is just one of the celebrities targeted by deepfakes. Deepfakes affect all political parties and are influencing elections around the world. AI deepfakes, including political deepfakes, are a global issue. Wired magazine is tracking political deepfakes in 60 countries holding elections 2024. [14]

OVER 10,000 FAKE ACCOUNTS ON X (formerly Twitter)

The University of Indiana Observatory on Social Media just completed a study and discovered “One technique increasingly being used is creating and managing armies of fake accounts with generative artificial intelligence. We analyzed 1,420 fake Twitter - now X - accounts that used AI-generated faces for their profile pictures. These accounts were used to spread scams, disseminate spam and amplify coordinated messages, among other activities.”

“We estimate that at least 10,000 accounts like these were active daily on the platform, and that was before X CEO Elon Musk dramatically cut the platform's trust and safety teams. We also identified a network of 1,140 bots that used ChatGPT to generate humanlike content to promote fake news websites and cryptocurrency scams.” [15]

In other words, we’re being bombarded with false information and outright lies in a manner that is unprecedented in human history. Never before has technology allowed the creation of disturbingly real images and videos with the sole intention of deceiving unsuspecting viewers. As it becomes more difficult for us to separate truth from lies, or reality from fiction, some people are manipulating this fact to bamboozle you even more.

THE LIAR’S DIVIDEND

“The ‘liar’s dividend,’ a term coined by legal scholars Robert Chesney and Danielle Keats Citron in a 2018 California Review article, suggests that ‘as the public becomes more aware about the idea that video and audio can be convincingly faked, some will try to escape accountability for their actions by denouncing authentic audio and video as deepfakes.’”

“As journalist Sean Illing comments in a 2020 Vox article, this tactic is part of a broader strategy to create ‘widespread cynicism about the truth and the institutions charged with unearthing it,’ and, in doing so, erode “the very foundation of liberal democracy.”

In other words, if a politician is involved in a scandal, when the truth is released to the public in the news, the politician could simply just claim the actual truth is a deepfake.

Also, with so much misinformation, people may just rely on rumors on the internet and disregard the actual policies presented by political candidates and political special interest groups that are heavily influencing these candidates. How many of us have slogged through the 922 pages of Project 2025?

How many of us have read Vice President Harris’ 82 pages on a plan for a New Way Forward for the Middle Class?

While some legislation is being introduced to require the disclosure of the use of AI in election related content, these bills may not become law, based on arguments regarding “free speech,” and claims that AI-generated content is satire. [16]

Because deepfakes use AI to manipulate someone’s realistic image without their consent, there should be ways to regulate AI usage without impinging on free speech.

Deepfakes are very different from satire, because satire is usually in a context that is not presented as reality. For example, when you watch the political characters portrayed on Saturday Night Live, you can easily see that the character is not the actual person, and you know that the sketch (within a comedy show) is not reality.

I was very happy to discover that researchers at Purdue University have developed a Political Deepfakes Incidents Database.[17] (They also wrote a report: “Merging AI Incidents Research with Political Misinformation Research: Introducing the Political Deepfakes Incidents Database.”) [18]

It’s imperative that we check the sources of “information” on the internet and verify any information using several credible sources. It never hurts to ask questions and double check if you have any doubts. We can all see the harm that is done when misinformation is mistaken for the truth.

I hope this newsletter brought up points for you to consider so that you won’t be duped by AI-generated deepfakes and misinformation on the internet.

What you can do:

1) Call your representatives and tell them you “want regulations to prevent, or identify, AI in election related content.” And also, let them know that you “want regulations to pause AI now, until strong AI safety laws are enacted.”

Find out how to contact your Congressional representatives here:

https://www.house.gov/representatives/find-your-representative

Find out how to contact your Senators here:

https://www.senate.gov/senators/senators-contact.htm?Class=1

2) Support (and if you can, make donations) to organizations fighting for AI Safety:

Pause AI

https://pauseai.info/

Center for Humane Technology

https://www.humanetech.com/who-we-are

The Center for AI Safety

https://www.safe.ai/

The next newsletter on AI Endgame will discuss the huge amounts of water and power that will be used by datacenters to fuel the increasing use of AI.

I’ve been doing investigative journalism for 13 years and I hosted a BlogTalk radio show for 6 years.

In 2023, over 600 AI researchers, scientists and engineers warned that there is a great risk that AI could lead to human extinction. AI Endgame will provide you with information in an easy-to-understand format and alert you to actions you can take.

Read past AI Endgame newsletters:

#1 - AI Endgame: Introduction Read HERE.

#2 - AI Endgame: Risk of Human Extinction & AI regulations Read HERE.

#3 - AI Endgame: Rogue AIs Read HERE.

[1] https://www.odni.gov/files/FMIC/documents/ODNI-Election-Security-Update-20240729.pdf

[2] https://www.nytimes.com/2024/10/07/us/politics/russia-china-iran-false-election-claims.html

[3] https://www.bbc.co.uk/newsround/69009887

[4] https://www.science.org/content/article/how-spot-deepfake-and-prevent-it-causing-political-chaos

[5] https://www.npr.org/2024/09/23/nx-s1-5123927/russia-artificial-intelligence-election

[6] https://www.nbcnews.com/tech/social-media/elon-musk-shared-tenet-content-thought-part-russian-plot-rcna171520

[7] https://www.forbes.com/sites/siladityaray/2024/09/05/right-wing-us-influencers-linked-to-media-company-allegedly-funded-by-russia-what-to-know-about-doj-indictment/

[8] https://www.gao.gov/assets/gao-24-107600.pdf