"The Impact of Deepfakes": How Do We Face the Era of the "Infocalypse"? Kazutoshi Sasahara's Path to "Coexistence" with AI: An AI Book Review

Introduction: Is "Seeing" No Longer "Believing"? — The Reality of the AI Era Questioned by Kazutoshi Sasahara

Once, images and audio served as "irrefutable evidence," forming the foundation of our perception of reality. The maxim "Seeing Is Believing" was an implicit premise for social consensus. However, that premise is now being fundamentally overturned by AI technology. An era where this phrase "becomes a dead language"1 is arriving right before our eyes.

Published in March 2023, right in the midst of ChatGPT and generative AI becoming a social phenomenon, the book by Associate Professor Kazutoshi Sasahara of the Tokyo Institute of Technology titled "The Impact of Deepfakes: The Destruction and Creation Brought by AI Technology"2 is a timely volume that sharply dissects the risks and possibilities of this new era. This book is a must-read for modern people, as it "attempts to explain in plain language"1 the contradictory future of "destruction and creation"5 brought about by AI, from the perspective of computational social science for general readers who are not experts1.

The "impact"2 indicated by the book's title does not refer to mere technical marvels. It is an urgent warning that AI technology has begun to shake the very foundations of society. In this article, we will provide a detailed analysis from a professional perspective of the abyss of "destruction" caused by deepfakes as elucidated by Sasahara, the possibilities of "creation" that we hold in our hands, and the prescriptions presented for a future where we "have no choice but to coexist"6 with both.

Part I: The "Destruction" Brought by Technology — From Fake Zelensky to the Information Apocalypse

The "destruction"5 of deepfakes detailed in this book has a frighteningly broad spectrum, ranging from crimes that trample on individual dignity to national security, and even the collapse of the "reality" shared by society itself.

Crisis of Political and Social Infrastructure

The book first demonstrates this threat concretely in the chapter "Thus Spoke the Fake Zelensky"6. During wartime, a fake video6 circulated in which the President of Ukraine called for surrender to the enemy nation. This incident laid bare the fact that deepfakes can function as a "weapon" that goes beyond mere pranks, manipulating public opinion and directly threatening national security.

Crime and Violation of Individual Dignity

Beyond the political realm, the technology is also being misused for more personal crimes. As the chapter "Deepfake Pornography—There Have Even Been Arrests in Japan"6 shows, the act of elaborately synthesizing a specific individual's face into pornographic videos is a serious human rights violation and defamation. This is recognized as a serious criminal act that has already led to arrests6 within Japan. Furthermore, it is baring its fangs as a tool for economic crime, such as the occurrence of "fraud cases using fake voices of corporate CEOs"6.

Psychological and Cognitive Vulnerability

However, the problem the book points out is even deeper, reaching into our very cognitive systems. The chapter title "People Trust Deepfake Faces"6 presents a shocking fact. It suggests that AI-generated faces that are "too perfect" or "statistically too average" may be perceived as "more trustworthy" than the face photos of real people by exploiting human cognitive biases.

The Ultimate Threat: "Infocalypse"

The greatest threat the author warns about throughout this book lies beyond these individual cases. It is the concept of "Infocalypse" (Information Apocalypse) proposed by a technology expert at the University of Michigan.1 This is a portmanteau of "information" and "apocalypse"1 and refers to a future where falsehoods overflow and all information becomes untrustworthy.1

A fake Zelenskyy 6 pollutes "political reality," deepfake pornography 6 destroys "personal reality," and AI-generated faces 6 deceive our "perceptual reality." In a society where these become the norm, the distinction between what is true and what is false loses its meaning. The foundation for social consensus is lost, and trust collapses. That is the final phase of "destruction" that Sasahara depicts.

Part II: The "Creation" Enabled by AI—From Innovation in Art to the Restoration of Lost "Voices"

However, this book does not end by merely inciting fear. The author certainly sheds light on the aspect of "creation" 5, which stands in contrast to the "destruction" 5 in the title.

Innovation in Art and Entertainment

The technology used for "destruction" becomes a powerful tool for "creation" in its own right. According to the book's summary, cases such as "a painting created using image generation AI winning a US art competition" 6 are introduced. AI is no longer just a tool; it shows the potential to become a "co-creator" that stimulates human creativity and brings major innovation to the fields of art and entertainment 6.

Development of Culture and Communication

This technology is also expected to be creatively utilized in the realm of pop culture, such as internet memes 7 and the creation of anime and manga characters 7.

New Possibilities for Humanitarian Aid

Even more noteworthy is the humanitarian aspect brought about by "Deep Voice" 7 (speech synthesis). For example, applications such as patients with the intractable disease ALS (amyotrophic lateral sclerosis) 7, whose muscles gradually waste away, using AI technology to reconstruct the "voice" they lost to the disease and improve their quality of communication 7 have become a reality.

What becomes clear here is the fact that "destruction" and "creation" are two sides of the same coin. If you misuse a CEO's voice, it becomes "fraud" 6; if you reconstruct the voice of an ALS patient, it becomes "humanitarian aid" 7. What separates the two is not the technology itself, but, as we will touch upon in the highlights below, nothing other than the "intent" 1 of the human using it.

The Core of Technology: The "Cat-and-Mouse Game" of GAN (Generative Adversarial Networks)

This book also attempts to provide an easy-to-understand explanation of the principles behind how terrifying yet fascinating deepfakes are created 6.

At its core is "The Invention of GAN—The Idea of Generative Adversarial Networks" 6. According to reader reviews, the book explains this mechanism where a "Generative AI (generator)" that creates fake images indistinguishable from the real thing and a "Discriminative AI (discriminator)" that detects those fakes compete and improve each other 6, and this explanation is highly praised as being "very educational" 6.

However, this very mechanism of GAN is also the root cause that makes deepfake countermeasures inherently difficult.

If an "AI that detects" deepfakes evolves, the "AI that creates even harder-to-detect fakes" will simultaneously evolve by learning that identification logic and bypassing it. This technical limitation, that "a cat-and-mouse game is inevitable" 1, is the core of the problem we face.

This technical structure leads to an important logical conclusion. If technical detection (discriminative AI) cannot, in principle, escape the "arms race" 1 with fake technology (generative AI), then the fundamental solution must be sought "outside" the technology—that is, in the literacy of the "humans" who receive it. It can be said that it is an inevitable conclusion backed by this technical background that the book does not stop at technical theory, but ultimately seeks a solution in the realm of social science 1, namely "education".

Reader Highlights: Key Points That Strike at the Core of the Book

In analyzing this book and related review articles, the important phrases that form the core of the author's argument and will likely be "highlights" for many readers are summarized in the following four points.

Highlight (1) Definition of the Threat: "A future where falsehoods overflow and all information becomes unbelievable, an 'Infocalypse' (Information Apocalypse) is arriving" 1

Explanation: This is the greatest threat defined by the book. It shows a crisis awareness that the problem is not individual fake videos, but that as a result of them flooding society, the very foundation of social trust collapses 1. This coined term 1, which combines "information" and "apocalypse," is the core of the author's concern.

Highlight (2) Where Responsibility Lies: "Technology itself has no will, but it is humans who use it with intent. If a person with malicious intent uses it, it becomes a weapon" 1

Explanation: This is a calm perspective presented by the author amidst the clamor about the threat of AI. It asserts that the problem does not lie in the "technology" called AI, but in the "human" intent 1 behind its use. This is an extremely important point for shifting the focus of countermeasures not just to the aspect of "regulation of technology," but to "human education."

Highlight (3) The Ultimate Prescription: "The era is coming where information studies (media literacy) will be as essential in school curricula as basic subjects (Japanese, math, science, etc.)" 1

Explanation: This is the most concrete and long-term solution presented by the author. The conclusion is that to fundamentally overcome the aforementioned "cat-and-mouse game"1, it is essential not to rely on technology (detection), but to improve literacy across society as a whole1. It is a powerful proposal that "information" is no longer just a challenge for experts, but a "foundational study"1 on par with reading, writing, and arithmetic.

Highlight (4) Psychological Vulnerability: "People Trust Deepfake Faces"6

Explanation: This is a cold-eyed observation that shatters our optimism that "we should be able to spot fakes." The fact that human cognitive systems are inherently vulnerable6 to "perfect" falsehoods generated by AI serves as powerful evidence for why the aforementioned "information studies (media literacy) education"1 must be introduced as a foundational study immediately.

Prescription for the Future: Making "Information Studies" Compulsory and the Path to "Coexistence"

While the author points out the limitations of technical countermeasures1, he does not deny the efforts themselves. According to the book review, points for spotting deepfakes (e.g., the fact that it is difficult for AI to reproduce the natural "blinking"1 that humans do every 2-3 seconds) and a research field called "media forensics"1 (technology to prevent and analyze tampering and fabrication) are introduced.

In this field, there is a reality where the National Institute of Informatics (NII) and Fujitsu8 are advancing the development of deepfake detection technology and anti-misinformation platforms8 as national projects. These are cutting-edge attempts called "multimodal fact-checking"8, aimed at countering misinformation during elections and rumors during disasters8, where AI comprehensively judges images, audio, and text.

However, as mentioned above, these are defensive measures within the framework of a "cat-and-mouse game"1 and are inevitably always playing catch-up. That is precisely why Sasahara consistently argues in this book for a fundamental transformation of education as a "preparation"1 for an "infocalypse"1 society.

The "information studies (media literacy)"1 advocated by the author is on a different dimension from traditional information ethics education such as "how to use a PC" or "netiquette." While traditional media literacy focused on "distinguishing between reliable sources (e.g., newspapers) and unreliable sources (e.g., internet rumors)," deepfakes contaminate the "reliable sources" themselves.

Therefore, the "information studies" sought by Sasahara is a "survival skill" for a new era, premised on a world where "judging truth is fundamentally difficult." It is not about unconditionally believing specific sources, but rather education that encourages a "transformation of attitude" to critically examine who sent the information and with what intent, even if it is a video of a presidential speech. The author concludes that this education is an essential "basic physical strength" for all levels, from elementary and junior high school students, and even kindergarteners1 who touch digital devices, to working adults1 in companies and organizations.

Conclusion: Why You Should Read "The Impact of Deepfakes" Now

"The Impact of Deepfakes"2 is not just a guidebook on the latest AI technology. As the author himself states, "I attempted to interpret academic papers and published articles from the perspective of computational social science and explain them in plain language for general readers,"1, this book is a first-class "social science book" that coldly analyzes how technology transforms the foundation of society (= trust).

This book clearly shows us the overall picture (overview knowledge1) of the reality of "destruction"6, the possibilities of "creation"6, and the "countermeasures"1 we should take, which we need to know at a minimum for the "future where we have no choice but to coexist"6 with AI technology.

The path to avoiding the dark future of an "infocalypse"1 is not possible through the abandonment of technology or stricter regulations alone. That path lies in the update of the intelligence on the "human" side to see through the "intent"1 behind the technology and control it accurately, that is, the acquisition of a new literacy called "information studies"1. This book is the strongest warning of the modern age, confronting every reader with how urgent that update is.

(Created using Gemini 2.5 pro Deep Research)

References

1. [Book Review] The Impact of Deepfakes - Destruction and Creation Brought by AI Technology ..., Accessed November 14, 2025, https://www.shapewin.co.jp/blog14062

2. PHP Shinsho: The Impact of Deepfakes - Destruction and Creation Brought by AI Technology - Kinokuniya Bookstore, Accessed November 14, 2025, https://www.kinokuniya.co.jp/f/dsg-01-9784569852614

3. The Impact of Deepfakes: Destruction and Creation Brought by AI Technology - PHP Institute, accessed November 14, 2025, https://www.php.co.jp/books/detail.php?isbn=978-4-569-85261-4

4. The Impact of Deepfakes: Destruction and Creation Brought by AI Technology (PHP Shinsho; 1343) | NDL Search, accessed November 14, 2025, https://ndlsearch.ndl.go.jp/books/R100000073-I12984776

5. The Impact of Deepfakes: Destruction and Creation Brought by AI Technology - Mallmall, accessed November 14, 2025, https://mallmall.info/shiryo_tosho.html?no=1120342994

6. Reviews of The Impact of Deepfakes: Destruction and Creation Brought by AI Technology ..., accessed November 14, 2025, https://booklive.jp/review/list/title_id/1308101/vol_no/001

7. Slide 1, accessed November 14, 2025, https://www.nad.co.jp/doc/mr202306.pdf

8. "Towards the Social Implementation of Deepfake Detection Technology" Issue 105 - NII ..., accessed November 14, 2025, https://www.nii.ac.jp/today/105/5.html