Deepfakes have become easier to deploy and more accurate, making them increasingly believable. This trend will likely continue, further blurring the line between real and manipulated media.
Once limited to entertainment, deepfakes now fuel political satire and misinformation. In recent years, deepfakes have mocked politicians or made them say outrageous things for entertainment. They often feature individuals like President Joe Biden, Vice President Kamala Harris, Governor Gavin Newsom, and President Donald Trump. While realistic, they still feature detectable markers, such as unnatural voice modulation, inconsistencies in mouth and face movements, and mismatched contexts–that easily reveal them as fakes. But what if these technologies were used in a more subtle and believable context? What if the stakes were much higher?
Deepfakes as a New Form of Warfare
As deepfakes grow in sophistication, their potential to disrupt national security becomes even more concerning. Their ability to deceive, disrupt, and influence at an unprecedented scale presents a significant threat. These technologies are not merely tools for entertainment; they have become a new weapon in modern warfare. Hostile actors now have the capability to disseminate misinformation among the American public, fueling polarization and distrust, and compromise the chain of command by impersonating government officials.
Fabricating Stories to Undermine Public Trust
One of the greatest dangers of deepfakes is the fabrication of false narratives with seemingly credible sources. Attackers can create deepfakes of fictional individuals falsely claiming that prominent leaders, such as the President, senators, or other high-ranking officials, did or said things they never actually did. These deepfakes can be used to introduce fictional assistants, low-ranking aides, or other seemingly credible figures claiming to have overheard confidential information or witnessed key events. Their perceived proximity to power makes their false claims appear believable.
Fabricated deepfake videos of fictional characters can quickly spread across social media, amplifying their impact and distorting political narratives. The viral nature of these platforms allows false claims to reach large audiences quickly, damaging officials’ reputations and eroding trust in institutions. By targeting fictional characters who seem "real," attackers can exploit cognitive biases, making people more likely to believe sensational stories that align with their existing views. This type of disinformation, fueled by the speed and reach of social media, could be incredibly effective in sowing confusion and mistrust on a global scale.
Real-Life Example: The Bugatti Deepfake
In June 2024, a deepfake video surfaced online featuring a non-existent employee of a Bugatti dealership in Paris. The video falsely claimed that Ukrainian President Volodymyr Zelensky’s wife had purchased a Bugatti Tourbillon for 4.5 million euros, complete with a fabricated invoice to enhance its credibility. Later investigations revealed that the video was part of a Russian disinformation campaign. This deepfake indirectly targeted a public figure by using a fabricated individual to lend false credibility. It demonstrates how attackers exploit both real and fictional characters to spread disinformation effectively.
Targeting High-Profile Figures to Manipulate Action
Attackers can target high-profile figures, such as the President or senior government officials, by impersonating them. These fabrications can deceive subordinates into taking actions that harm national interests. Misled by fake messages from trusted sources, decision-makers may unknowingly implement policies or take actions that serve hostile interests. Such deceptions can lead subordinates to act in ways that benefit malicious actors, ultimately jeopardizing U.S. national security and political stability.
Real-Life Example: Fake AI Robocalls Targeting Voters
One recent incident demonstrates how easily deepfakes can manipulate public perception and influence democratic processes. In January 2024, a political consultant orchestrated AI-generated robocalls impersonating President Biden’s voice, calling voters in New Hampshire ahead of the presidential primary. The calls misled voters into thinking that participating in the primary would prevent them from voting in the November election. Although the deception was quickly detected, it still reached many Americans and disrupted the political process.
Real-Life Example: Targeted Deepfake Attack on U.S. Senator
In another case, U.S. Senator Ben Cardin became a target of a deepfake attack in September 2024. Attackers used live deepfake technology to impersonate Ukraine's former Foreign Affairs Minister. Cardin and his staff grew suspicious after receiving unusual questions, prompting them to end the call. The U.S. State Department later confirmed that the caller was not the minister. Had the attack succeeded, it could have had serious ramifications for U.S. foreign policy, undermining the integrity of our government processes and strategy.
Potential Scenarios in 2025
Although previous deepfake incidents were contained quickly, they exposed vulnerabilities that could lead to more disruptive attacks. If malicious actors infiltrate government channels using deepfakes to impersonate high-level officials, the consequences could be especially disastrous. Here are some potential scenarios that could unfold this year:
Public Panic
Deepfakes could create realistic media of fictitious events, such as a fake terrorist attack or natural disaster, potentially inciting chaos, overwhelming emergency systems, and distorting news coverage–making fact indistinguishable from fiction.
First Responders and Military
Real-time deepfake technology could also be used in swatting attacks, where criminals impersonate victims’ voices during a crisis, such as a hostage situation, to manipulate law enforcement or individuals into acting. These attacks could divert critical resources, overwhelming emergency services with fake emergencies while real threats go unnoticed. In military settings, adversaries could use deepfakes to spread false information or impersonate key figures, disrupting intelligence agencies, misdirecting military resources, or even sabotaging operations. Such tactics could weaken national defense and first responder capabilities by spreading confusion and diverting attention from legitimate threats.
Critical Infrastructure
Attackers could use deepfakes to impersonate emergency operators or infrastructure personnel, issuing fake instructions that disrupt essential services such as energy, water, and healthcare. These false directives could destabilize national security and fuel public panic.
What Now?
Deepfakes can be entertaining in contexts like memes, video games, and the entertainment industry. However, in the wrong hands, manipulated media can spread misinformation, erode public trust, threaten national security, and undermine democracy. Deepfakes have become a sophisticated weapon in modern warfare. Whether through mass disinformation campaigns or the targeted manipulation of high-ranking officials, they pose immense risks.
Given the growing threat of deepfake manipulation, advanced assessment tools like Authentica are pivotal in identifying and attributing such fabrications so that we can distinguish malicious manipulated media. It’s not enough to detect whether a video, image, or audio file is fake. Authentica goes further–determining “Who” created the deepfake and “How” it was made, so our customers can answer the question “What Now.” As the new year progresses, it’s essential for government officials to acquire the necessary tools to combat this new age threat. Click here for more information on how to get access to Authentica and receive a free demo.
What are your thoughts? Join the conversation and leave a comment below!