In a new study examining the ways in which malicious actors deploy deepfakes to disinform, confuse, and defraud, Google’s researchers reiterated a grim prediction about generative AI and deepfakes having the potential to “distort collective understanding of socio-political reality or scientific consensus.” In other words, the flood of AI-generated images, text, audio, and video has the potential to subvert the pillars of reality most of us have come to agree upon, regardless of our ideological differences.
In the words of Mark Twain, “a lie can travel halfway around the world while the truth is still putting on its shoes.” A quote many of us admire — will we like it just the same in five years, after a deepfake of a maligned and long-deceased public figure utters the words from his synthetic lips in a viral video? The power of a lie isn’t only in the speed at which it travels, but in its instinctual impact. A lie is salacious, touches upon the primal, it provokes in us an emotional response where a truth tends to be bland and difficult at the same time, like eating a bowl of unseasoned beans.
The Heightened Importance of Reality
The Internet has given us remarkable ways to share information in seconds, changing the world at a pace that was once unimaginable. But its sinister side is more exploitable than ever thanks to the proliferation of AI-generated content. Digital spaces are overrun with mechanisms that make falsehoods travel faster than truths. Social media algorithms reward provocation over information, and engagement farmers race to exploit this trend.
The business model of platforms aimed at making us comfortable, or comfortably outraged, exploits the impulses all of us have, regardless of ideology: we want to have our beliefs confirmed, our biases appeased. Wherever appeasement cannot be achieved, chaos will do. Generative AI and deepfakes are simply the newest tools that make the distortion of reality to make it seem like our own, customized reality, easier to achieve. What’s terrifying is how effective they are at doing just that, while the technology is still in its infancy.
But a truth customized to each individual based on the subscription or product a company wants to sell cannot sustain the social cohesion we need for humanity to function. While dissent is necessary and healthy, shared truths and realities hold us together — it is hard to deny that our agreements to make decisions based on evidence and form representative systems of government have propelled many societies toward a brighter future.
Distorted Reality and Disinformation
The distortion of reality is, of course, even more concerning in the realm of politics, where the distortion of facts can lead to irreversible consequences that reach far beyond social cohesion. A recent World Economic Forum report declared deepfake-based disinformation as the likeliest reason for immediate destabilization of global society.
As AI and deepfake technology advance, we risk losing another crucial domain in our fight to preserve our sense of reality—we will no longer be able to trust our eyes and ears. To counter these emerging threats, we must strive harder than ever to protect our shared reality. Humanity thrives by inventing effective responses to threats. AI, while creating the problem, also offers solutions by powering advanced deepfake detection models that can flag deepfakes in real-time, despite the overwhelming flood of AI-generated content.
Organizations are educating the public about disinformation efforts designed to distort their sense of truth, and digital platforms globally are seeking responsible ways to verify information before it reaches users. While there are plenty of nefarious actors using AI for their own purposes, many of us are banding together to do what is necessary to resist these efforts.
The good news is that we know what tools are needed to preserve our sense of truth. By spreading public awareness and supporting media and AI literacy, bolstering fact-checking efforts across platforms and industries, and constantly improving deepfake detection methods, we can continue to strengthen our commitment to actively defend reality.
