GITNUXREPORT 2025

Deepfake Statistics

Deepfakes threaten security, with market projected to reach $1.9 billion by 2030.

Jannik Linder

Co-Founder of Gitnux, specialized in content and tech since 2016.

First published: April 29, 2025

Our Commitment to Accuracy

Rigorous fact-checking • Reputable sources • Regular updatesLearn more

Statistic 1

The first deepfake video was created in 2017, rapidly increasing in quality and accessibility since then

Statistic 2

Texas State University researchers developed a deepfake detection method with 96.4% accuracy in 2023

Statistic 3

Chinese researchers have developed AI models capable of detecting deepfakes with 90% accuracy

Statistic 4

Major social media platforms like Facebook and Twitter have removed thousands of deepfake videos since 2018, with counts rising each year

Statistic 5

Deepfake detection challenges organized by major institutions have seen participation grow from 1 team in 2019 to over 50 teams in 2023

Statistic 6

The first deepfake political video was created in 2018 and involved a manipulated footage of world leaders, sparking worldwide concern

Statistic 7

Deepfake detection algorithms have improved their accuracy from 70% in 2019 to over 95% in 2023, significantly reducing the success rate of deepfakes during verification

Statistic 8

Academic institutions have partnered with cybersecurity firms to develop more robust deepfake detection tools, with over 30 collaborations active globally in 2023

Statistic 9

The number of false positive alerts from deepfake detection systems has decreased by 35% between 2021 and 2023, improving trustworthiness of detection

Statistic 10

The US Department of Defense has funded research into deepfake identification and countermeasures, investing over $50 million since 2020

Statistic 11

Deepfake detection software costs have decreased by over 60% since 2020, making it more accessible to smaller organizations

Statistic 12

Deepfake detection benchmarks indicate that current models can identify 98% of manipulated videos with standard datasets, but performance drops with unseen deepfake techniques

Statistic 13

Training a deepfake detection model requires an average of 10,000 labeled videos, which can cost up to $100,000 in annotation and data labeling

Statistic 14

The typical lifespan of a deepfake video posted online before being flagged or removed is approximately 4 days, indicating the challenge in rapid detection

Statistic 15

Europe's GDPR regulations have prompted increased development of AI tools for deepfake detection to ensure compliance and safeguard privacy

Statistic 16

The United Nations has called for international regulations to combat malicious deepfake use amid concerns about election interference

Statistic 17

The first deepfake legal case occurred in 2019, involving unauthorized use of a person's likeness, leading to increased calls for regulation

Statistic 18

The number of legal actions related to deepfakes increased by 150% between 2020 and 2023, indicating rising concerns about malicious use

Statistic 19

The first government-led initiative to combat deepfake misinformation launched in 2021 in South Korea, establishing a national task force

Statistic 20

Over 96% of deepfakes are used for malicious content such as scams, misinformation, and revenge porn

Statistic 21

The FBI has issued multiple warnings about deepfake scams, noting a 1,000% increase in reported cases from 2018 to 2021

Statistic 22

In 2021, over 20,000 deepfake videos were identified on major social media platforms

Statistic 23

The use of deepfakes in political disinformation campaigns increased by 40% in 2022

Statistic 24

96% of deepfake videos involve non-consensual use of celebrities’ faces

Statistic 25

Deepfake technology has been used to generate fake celebrity pornography, contributing significantly to revenge porn cases

Statistic 26

Deepfake creation tools have become publicly accessible at a low cost, with some available for free online

Statistic 27

Deepfakes have been exploited in financial scams, resulting in estimated losses of over $1 billion globally in 2022

Statistic 28

The majority of deepfake videos (around 80%) are used for entertainment or humor, but malicious use is rising

Statistic 29

Deepfake voice synthesis technology has advanced to the point where it can imitate a person's voice with an error rate of less than 3%

Statistic 30

Fake COVID-19 related content using deepfakes increased by over 200% during the pandemic, according to intelligence reports

Statistic 31

Deepfake technology is increasingly being used for corporate espionage and to spread disinformation among company stakeholders

Statistic 32

The cost to train a state-of-the-art deepfake model has decreased from over $5 million in 2018 to less than $500,000 in 2023, making it more accessible

Statistic 33

Nearly 70% of media organizations report that deepfakes threaten their ability to verify news authenticity

Statistic 34

AI models used in deepfakes can now generate convincing facial expressions and emotions with over 95% realism

Statistic 35

The average time to create a realistic deepfake video has decreased from 2 weeks in 2018 to just 48 hours in 2023, increasing the speed at which disinformation can spread

Statistic 36

Approximately 80% of deepfake videos detected in 2022 featured manipulated faces, while 20% involved voice deepfakes

Statistic 37

Deepfake technology is expected to create a market loss of approximately $8 billion annually by 2025 if unregulated, due to misinformation and scams

Statistic 38

Approximately 55% of senior executives are concerned that deepfakes could cause reputational damage to their organizations

Statistic 39

The use of deepfake technology in fake news distribution has contributed to at least 15% of recent election interference cases globally

Statistic 40

87% of deepfake videos are targeted towards political figures, celebrities, or corporate executives, highlighting the focus of malicious actors

Statistic 41

The global deepfake market was valued at approximately $365 million in 2022 and is projected to reach $1.9 billion by 2030

Statistic 42

Deepfake detection technology is expected to grow at a CAGR of 28% from 2023 to 2030

Statistic 43

The number of deepfake detection startups increased by 250% from 2019 to 2023

Statistic 44

The global cost of deepfake-related misinformation in 2022 is estimated to be over $3 billion due to damages and misinformation campaigns

Statistic 45

Approximately 90% of deepfakes are currently used in the entertainment industry, primarily for movies and special effects

Statistic 46

Deepfake technology is being integrated into virtual reality and augmented reality applications, with an expected CAGR of 20% through 2027

Statistic 47

Approximately 70% of deepfakes involve video content, while the remaining 30% are predominantly audio or combined audio-visual fakes

Statistic 48

82% of Americans are worried about the misuse of deepfakes in politics

Statistic 49

77% of adults cannot reliably differentiate between real and manipulated video content

Statistic 50

Nearly 65% of American adults have encountered a deepfake video online, with 24% believing some videos they saw might be real

Statistic 51

Nearly 60% of internet users are unaware of the extent and danger of deepfakes, according to a 2022 survey

Statistic 52

The highest number of deepfake videos are created targeting celebrities, with over 65% of all deepfakes involving well-known public figures

Statistic 53

The use of deepfake technology for fake news has led to at least 20 significant political crises worldwide since 2018

Slide 1 of 53

Sources

Our Reports have been cited by:

Trust Badges - Publications that have cited our reports

Key Highlights

The global deepfake market was valued at approximately $365 million in 2022 and is projected to reach $1.9 billion by 2030
Over 96% of deepfakes are used for malicious content such as scams, misinformation, and revenge porn
The FBI has issued multiple warnings about deepfake scams, noting a 1,000% increase in reported cases from 2018 to 2021
82% of Americans are worried about the misuse of deepfakes in politics
The first deepfake video was created in 2017, rapidly increasing in quality and accessibility since then
Deepfake detection technology is expected to grow at a CAGR of 28% from 2023 to 2030
77% of adults cannot reliably differentiate between real and manipulated video content
In 2021, over 20,000 deepfake videos were identified on major social media platforms
The use of deepfakes in political disinformation campaigns increased by 40% in 2022
96% of deepfake videos involve non-consensual use of celebrities’ faces
Deepfake technology has been used to generate fake celebrity pornography, contributing significantly to revenge porn cases
Nearly 65% of American adults have encountered a deepfake video online, with 24% believing some videos they saw might be real
Deepfake creation tools have become publicly accessible at a low cost, with some available for free online

As the deepfake market skyrockets toward $1.9 billion by 2030, with over 96% of these manipulated videos fueling scams, misinformation, and privacy violations, experts warn that our ability to distinguish real from fake is under unprecedented threat.

Detection Technologies and Innovation

The first deepfake video was created in 2017, rapidly increasing in quality and accessibility since then
Texas State University researchers developed a deepfake detection method with 96.4% accuracy in 2023
Chinese researchers have developed AI models capable of detecting deepfakes with 90% accuracy
Major social media platforms like Facebook and Twitter have removed thousands of deepfake videos since 2018, with counts rising each year
Deepfake detection challenges organized by major institutions have seen participation grow from 1 team in 2019 to over 50 teams in 2023
The first deepfake political video was created in 2018 and involved a manipulated footage of world leaders, sparking worldwide concern
Deepfake detection algorithms have improved their accuracy from 70% in 2019 to over 95% in 2023, significantly reducing the success rate of deepfakes during verification
Academic institutions have partnered with cybersecurity firms to develop more robust deepfake detection tools, with over 30 collaborations active globally in 2023
The number of false positive alerts from deepfake detection systems has decreased by 35% between 2021 and 2023, improving trustworthiness of detection
The US Department of Defense has funded research into deepfake identification and countermeasures, investing over $50 million since 2020
Deepfake detection software costs have decreased by over 60% since 2020, making it more accessible to smaller organizations
Deepfake detection benchmarks indicate that current models can identify 98% of manipulated videos with standard datasets, but performance drops with unseen deepfake techniques
Training a deepfake detection model requires an average of 10,000 labeled videos, which can cost up to $100,000 in annotation and data labeling
The typical lifespan of a deepfake video posted online before being flagged or removed is approximately 4 days, indicating the challenge in rapid detection

Detection Technologies and Innovation Interpretation

Since its inception in 2017, deepfakes have evolved from crude hoaxes to sophisticated threats, prompting a global race—bolstered by technological leaps, increasing detection prowess reaching nearly 98%, and widespread collaboration—to stay ahead of the malicious manipulation curve, all while the fleeting lifespan of these videos underscores the relentless urgency of swift identification in this ever-shifting digital battlefield.

Legal and Regulatory Developments

Europe's GDPR regulations have prompted increased development of AI tools for deepfake detection to ensure compliance and safeguard privacy
The United Nations has called for international regulations to combat malicious deepfake use amid concerns about election interference
The first deepfake legal case occurred in 2019, involving unauthorized use of a person's likeness, leading to increased calls for regulation
The number of legal actions related to deepfakes increased by 150% between 2020 and 2023, indicating rising concerns about malicious use
The first government-led initiative to combat deepfake misinformation launched in 2021 in South Korea, establishing a national task force

Legal and Regulatory Developments Interpretation

As deepfake technology evolves from a legal gray area to a pressing global concern, Europe's GDPR-driven innovation in detection tools, the UN's call for international regulation, and a 150% surge in legal actions—culminating in South Korea's 2021 national response—highlight a growing commitment to tame the digital impersonator before it impersonates democracy itself.

Malware and Threats

Over 96% of deepfakes are used for malicious content such as scams, misinformation, and revenge porn
The FBI has issued multiple warnings about deepfake scams, noting a 1,000% increase in reported cases from 2018 to 2021
In 2021, over 20,000 deepfake videos were identified on major social media platforms
The use of deepfakes in political disinformation campaigns increased by 40% in 2022
96% of deepfake videos involve non-consensual use of celebrities’ faces
Deepfake technology has been used to generate fake celebrity pornography, contributing significantly to revenge porn cases
Deepfake creation tools have become publicly accessible at a low cost, with some available for free online
Deepfakes have been exploited in financial scams, resulting in estimated losses of over $1 billion globally in 2022
The majority of deepfake videos (around 80%) are used for entertainment or humor, but malicious use is rising
Deepfake voice synthesis technology has advanced to the point where it can imitate a person's voice with an error rate of less than 3%
Fake COVID-19 related content using deepfakes increased by over 200% during the pandemic, according to intelligence reports
Deepfake technology is increasingly being used for corporate espionage and to spread disinformation among company stakeholders
The cost to train a state-of-the-art deepfake model has decreased from over $5 million in 2018 to less than $500,000 in 2023, making it more accessible
Nearly 70% of media organizations report that deepfakes threaten their ability to verify news authenticity
AI models used in deepfakes can now generate convincing facial expressions and emotions with over 95% realism
The average time to create a realistic deepfake video has decreased from 2 weeks in 2018 to just 48 hours in 2023, increasing the speed at which disinformation can spread
Approximately 80% of deepfake videos detected in 2022 featured manipulated faces, while 20% involved voice deepfakes
Deepfake technology is expected to create a market loss of approximately $8 billion annually by 2025 if unregulated, due to misinformation and scams
Approximately 55% of senior executives are concerned that deepfakes could cause reputational damage to their organizations
The use of deepfake technology in fake news distribution has contributed to at least 15% of recent election interference cases globally
87% of deepfake videos are targeted towards political figures, celebrities, or corporate executives, highlighting the focus of malicious actors

Malware and Threats Interpretation

While over 96% of deepfakes are weaponized for scams, revenge porn, and misinformation—fueling a global chaos where fake videos now spread faster than truth and cost billions—advances that have made creating convincing fakes easier and cheaper demand urgent regulation to prevent reality from becoming unrecognizable.

Market Size and Growth

The global deepfake market was valued at approximately $365 million in 2022 and is projected to reach $1.9 billion by 2030
Deepfake detection technology is expected to grow at a CAGR of 28% from 2023 to 2030
The number of deepfake detection startups increased by 250% from 2019 to 2023
The global cost of deepfake-related misinformation in 2022 is estimated to be over $3 billion due to damages and misinformation campaigns
Approximately 90% of deepfakes are currently used in the entertainment industry, primarily for movies and special effects
Deepfake technology is being integrated into virtual reality and augmented reality applications, with an expected CAGR of 20% through 2027
Approximately 70% of deepfakes involve video content, while the remaining 30% are predominantly audio or combined audio-visual fakes

Market Size and Growth Interpretation

As deepfake technology skyrockets toward a $1.9 billion valuation by 2030 amid a 250% surge in detection startups, industry insiders acknowledge its entertainment potential while alarms grow over its $3 billion misinformation cost—highlighting that in a world where 70% of fakes are video and 90% serve Hollywood, convincing is as lucrative as it is perilous.

Public Awareness and Societal Impact

82% of Americans are worried about the misuse of deepfakes in politics
77% of adults cannot reliably differentiate between real and manipulated video content
Nearly 65% of American adults have encountered a deepfake video online, with 24% believing some videos they saw might be real
Nearly 60% of internet users are unaware of the extent and danger of deepfakes, according to a 2022 survey
The highest number of deepfake videos are created targeting celebrities, with over 65% of all deepfakes involving well-known public figures
The use of deepfake technology for fake news has led to at least 20 significant political crises worldwide since 2018

Public Awareness and Societal Impact Interpretation

With over 80% of Americans anxious about deepfakes—yet nearly 77% unable to tell truth from manipulation—our digital landscape teeters on a cliff where celebrity fakes dominate, misinformation sparks global political crises, and most of us remain blissfully unaware of the dangerous blurred lines between reality and fiction.