Deepfake technology, AI‑driven synthetic media that can swap faces, mimic voices, or alter images, has leapt from niche experiment to global threat. Organizations now see deepfakes used in financial fraud, identity attacks, and political manipulation, while media platforms must scramble to detect manipulated content in real time. As the volume...

Introduction

DeepFake AI Statistics: DeepFake technology, fueled by cutting-edge Artificial Intelligence (AI), has quickly gained attention for its ability to create highly convincing alterations in audio and visual content. Through the use of deep learning algorithms, DeepFakes can modify videos, images, and voice recordings to make them appear as though individuals are saying or doing things they never actually did.

Initially developed for entertainment and creative applications, this technology has raised significant ethical issues, particularly in the realms of misinformation, privacy violations, and political manipulation. As DeepFake technology becomes more accessible, detecting these manipulations is becoming increasingly difficult, leading to growing concerns about trust in digital media.

In light of these challenges, it is crucial to understand the statistics and trends related to DeepFake AI to better address its implications across various sectors, including social media, law enforcement, and cybersecurity...

According to a survey conducted in March 2025, ** percent of adult female respondents in the United States expressed concerns about the spread of artificial intelligence (AI) video and audio deepfakes. Similarly, nearly ** percent of men shared this concern. In contrast, only *** percent of adult women and *** percent of adult men in the U.S. reported that they were not concerned at all.

Between February 2024 and February 2025, around 71 percent of surveyed employees in worldwide organizations stated being very or extremely confident in recognizing deepfake identity documents, such as ID cards or passports. However, the confidence in recognizing audio deepfakes was lower, with nearly 69 percent of employees stating so. Nevertheless, approximately 74 percent of the surveyed security officers stated the same.

This study conducts research on deepfakes technology evolution and trends based on a bibliometric analysis of the articles published on this topic along with six research questions: What are the main research areas of the articles in deepfakes? What are the main current topics in deepfakes research and how are they related? Which are the trends in deepfakes research? How do topics in deepfakes research change over time? Who is researching deepfakes? Who is funding deepfakes research? We have found a total of 331 research articles about deepfakes in an analysis carried out on the Web of Science and Scopus databases. This data serves to provide a complete overview of deepfakes. Main insights include: different areas in which deepfakes research is being performed; which areas are the emerging ones, those that are considered basic, and those that currently have the most potential for development; most studied topics on deepfakes research, including the different artificial intelligence methods applied; emerging and niche topics; relationships among the most prominent researchers; the countries where deepfakes research is performed; main funding institutions. This paper identifies the current trends and opportunities in deepfakes research for practitioners and researchers who want to get into this topic.

In January 2025, a small fintech startup in Austin discovered it had fallen victim to a cyberattack. At first glance, the breach looked like a typical case of credential stuffing. But it wasn’t. The attacker had used an AI-driven system that mimicked the behavioral patterns of employees, learning login habits,...

DeepFake Videos for detection tasks

Dataset consists of 10,000+ files featuring 7,000+ people, providing a comprehensive resource for research in deepfake detection and deepfake technology. It includes real videos of individuals with AI-generated faces overlaid, specifically designed to enhance liveness detection systems. By utilizing this dataset, researchers can advance their understanding of deepfake generation and improve the performance of detection methods. - Get the data… See the full description on the dataset page: https://huggingface.co/datasets/UniDataPro/deepfake-videos-dataset.

Description

SWAN-DF: the first high fidelity publicly available dataset of realistic audio-visual deepfakes, where both faces and voices appear and sound like the target person. The SWAN-DF dataset is based on the public SWAN database of real videos recorded in HD on iPhone and iPad Pro (in year 2019). For 30 pairs of manually selected people from SWAN, we swapped faces and voices using several autoencoder-based face swapping models and using several blending techniques from the well-known open source repo DeepFaceLab and voice conversion (or voice cloning) methods, including zero-shot YourTTS, DiffVC, HiFiVC, and several models from FreeVC.

For each model and each blending technique, there are 960 video deepfakes. We used three types of models of the following resolutions: 160x160, 256x256, and 320x320 pixels. We took one pre-trained model corresponding for each resolution, and tuned it for each of the 30 pairs (both ways) of subjects for 50K iterations. Then, when generating deepfake videos for each pair of subjects, we used one of the tuned models and a way to blend the generated image back into the original frame, which we call blending technique. SWAN-DF dataset contains 25 different combinations of models and blending, which means the total number of deepfake videos is 960*25=24000.

We generated speech deepfakes using four voice conversion methods: YourTTS, HiFiVC, DiffVC, and FreeVC. We did not use text to speech methods for our video deepfakes, since the speech they produce is not synchronized with the lip movements in the video. For YourTTS, HiFiVC, and DiffVC methods, we used the pretrained models provided by the authors. HiFiVC was pretrained on VCTK, DiffVC on LibriTTS, and YourTTS on both VCTK and LibriTTS datasets. For FreeVC, we generated audio deepfakes for several variants: using the provided pretrained models (for 16Hz with and without pretrained speaker encoder and for 24Hz with pretrained speaker encoder) as is and by tuning 16Hz model either from scratch or starting from the pretrained version for different number of iterations on the mixture of VCTK and SWAN data. In total, SWAN-DF contains 12 different variations of audio deepfakes: one for each of YourTTS, HiFiVC, and DiffVC and 9 variants of FreeVC.

Acknowledgements

If you use this database, please cite the following publication:

Pavel Korshunov, Haolin Chen, Philip N. Garner, and Sébastien Marcel, "Vulnerability of Automatic Identity Recognition to Audio-Visual Deepfakes", IEEE International Joint Conference on Biometrics (IJCB), September 2023. https://publications.idiap.ch/publications/show/5092

'In-the-Wild' Dataset We present a dataset of audio deepfakes (and corresponding benign audio) for a set of politicians and other public figures, collected from publicly available sources such as social networks and video streaming platforms. For n = 58 celebrities and politicians, we collect both bona-fide and spoofed audio. In total, we collect 20.8 hours of bona-fide and 17.2 hours of spoofed audio. On average, there are 23 minutes of bona-fide and 18 minutes of spoofed audio per speaker.

The dataset is intended to be used for evaluating deepfake detection and voice anti-spoofing machine-learning models. It is especially useful to judge a model's capability to generalize to realistic, in-the-wild audio samples. Find more information in our paper, and download the dataset here.

The most interesting deepfake detection models we used in our experiments are open-source on GitHub:

RawNet 2 RawGAT-ST PC-Darts This dataset and the associated documentation are licensed under the Apache License, Version 2.0.

A May 2024 survey of IT and cybersecurity professionals worldwide found that around ** percent of organizations in the United States were targeted by AI-assisted deepfakes in the past ** months, compared to the global average of ** percent.

Deepfake Image Dataset Description

The Deepfake Image Dataset is a comprehensive resource curated for researchers, developers, and data scientists working on detecting, analyzing, and understanding deepfakes. This dataset is meticulously structured to support machine learning and artificial intelligence applications, particularly in enhancing deepfake detection systems.

The dataset comprises high-quality images stored in zip file format, ensuring ease of access and efficient download. It is divided into two primary subsets: Training Data and Testing Data, enabling seamless development and evaluation of detection models.

• Training Dataset:
  This section contains hundreds of labeled images, including both authentic and deepfake-generated visuals. The images cover diverse scenarios, facial expressions, and environments to provide a robust foundation for model training. Each image is labeled with metadata specifying its category, ensuring straightforward integration with machine learning pipelines.

• Testing Dataset:
  Designed for performance validation, the testing dataset mirrors the complexity and diversity of the training set. It ensures reliable benchmarking of detection algorithms, facilitating accurate evaluation across varied conditions.

The dataset includes deepfake images generated using state-of-the-art techniques, reflecting real-world deepfake challenges. Each zip file is organized for quick unpacking and usability, with a consistent naming convention and directory structure for hassle-free navigation.

Ideal for applications in cybersecurity, digital forensics, and AI ethics, this dataset is a vital tool in combating the misuse of deepfake technology. By providing a reliable platform for experimentation and innovation, it empowers the global community to enhance digital integrity.

The main purpose of this data set is to facilitate research into audio DeepFakes. We hope that this work helps in finding new detection methods to prevent such attempts. These generated media files have been increasingly used to commit impersonation attempts or online harassment.

The data set consists of 104,885 generated audio clips (16-bit PCM wav). We examine multiple networks trained on two reference data sets. First, the LJSpeech data set consisting of 13,100 short audio clips (on average 6 seconds each; roughly 24 hours total) read by a female speaker. It features passages from 7 non-fiction books and the audio was recorded on a MacBook Pro microphone. Second, we include samples based on the JSUT data set, specifically, basic5000 corpus. This corpus consists of 5,000 sentences covering all basic kanji of the Japanese language (4.8 seconds on average; roughly 6.7 hours total). The recordings were performed by a female native Japanese speaker in an anechoic room. Finally, we include samples from a full text-to-speech pipeline (16,283 phrases; 3.8s on average; roughly 17.5 hours total). Thus, our data set consists of approximately 175 hours of generated audio files in total. Note that we do not redistribute the reference data.

We included a range of architectures in our data set:

MelGAN

Parallel WaveGAN

Multi-Band MelGAN

Full-Band MelGAN

WaveGlow

Additionally, we examined a bigger version of MelGAN and include samples from a full TTS-pipeline consisting of a conformer and parallel WaveGAN model.

Collection Process

For WaveGlow, we utilize the official implementation (commit 8afb643) in conjunction with the official pre-trained network on PyTorch Hub. We use a popular implementation available on GitHub (commit 12c677e) for the remaining networks. The repository also offers pre-trained models. We used the pre-trained networks to generate samples that are similar to their respective training distributions, LJ Speech and JSUT. When sampling the data set, we first extract Mel spectrograms from the original audio files, using the pre-processing scripts of the corresponding repositories. We then feed these Mel spectrograms to the respective models to obtain the data set. For sampling the full TTS results, we use the ESPnet project. To make sure the generated phrases do not overlap with the training set, we downloaded the common voices data set and extracted 16.285 phrases from it.

This data set is licensed with a CC-BY-SA 4.0 license.

This work was supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany's Excellence Strategy -- EXC-2092 CaSa -- 390781972.

The Deepfake AI market is experiencing rapid growth, driven by advancements in artificial intelligence and machine learning. While precise market sizing data is unavailable, considering the burgeoning interest in AI-powered media manipulation and its applications across various sectors, a reasonable estimate for the 2025 market size could be around $500 million. This is based on the increasing adoption of deepfake technology in entertainment, marketing, and education, as well as its potential in security and defense applications. The Compound Annual Growth Rate (CAGR) is projected to remain strong, potentially exceeding 25% through 2033, indicating significant market expansion. Key drivers include the decreasing cost and increasing accessibility of deepfake creation tools, coupled with rising demand for realistic video and audio content across various industries. However, ethical concerns surrounding misuse, such as the creation of fraudulent content and identity theft, present significant restraints to market growth. Regulatory frameworks and technological solutions aimed at detecting and mitigating deepfakes are crucial for fostering responsible innovation within the sector. The market is segmented by application (entertainment, advertising, security, etc.) and technology (generative adversarial networks (GANs), autoencoders, etc.). The competitive landscape is marked by a diverse range of companies offering both deepfake creation and detection solutions. Major players include Synthesia, Pindrop, Reface, and others, each focusing on specific niches within the market. The future of the Deepfake AI market hinges on the balance between technological advancement, ethical considerations, and robust regulatory oversight. Further growth will be significantly impacted by developments in AI detection technologies, increased public awareness, and clear industry guidelines aimed at minimizing malicious applications while encouraging beneficial uses. The geographical distribution is expected to show a strong concentration in North America and Europe initially, but significant growth potential exists in other regions as technology adoption matures globally. Furthermore, advancements in areas like AI-driven content verification and authentication are crucial for mitigating the risks associated with deepfake technology and unlocking its wider potential.

Small-scale Deepfake Forgery Video Dataset (SDFVD) is a custom dataset consisting of real and deepfake videos with diverse contexts designed to study and benchmark deepfake detection algorithms. The dataset comprising of a total of 106 videos, with 53 original and 53 deepfake videos. Equal number of real and deepfake videos, ensures balance for machine learning model training and evaluation. The original videos were collected from Pexels: a well- known provider of stock photography and stock footage(video). These videos include a variety of backgrounds, and the subjects represent different genders and ages, reflecting a diverse range of scenarios. The input videos have been pre-processed by cropping them to a length of approximately 4 to 5 seconds and resizing them to 720p resolution, ensuring a consistent and uniform format across the dataset. Deepfake videos were generated using Remaker AI employing face-swapping techniques. Remaker AI is an AI-powered platform that can generate images, swap faces in photos and videos, and edit content. The source face photos for these swaps were taken from Freepik: is an image bank website provides contents such as photographs, illustrations and vector images. SDFVD was created due to the lack of availability of any such comparable small-scale deepfake video datasets. Key benefits of such datasets are: • In educational settings or smaller research labs, smaller datasets can be particularly useful as they require fewer resources, allowing students and researchers to conduct experiments with limited budgets and computational resources. • Researchers can use small-scale datasets to quickly prototype new ideas, test concepts, and refine algorithms before scaling up to larger datasets. Overall, SDFVD offers a compact but diverse collection of real and deepfake videos, suitable for a variety of applications, including research, security, and education. It serves as a valuable resource for exploring the rapidly evolving field of deepfake technology and its impact on society.

In 2024, Poles learned false information generated by artificial intelligence most often through facial features and, above all, movement inconsistent with spoken words.

According to a survey conducted in March 2025 in the United States, ** percent of respondents aged 65 and older in the United States reported being very concerned about the spread of video and audio deep fakes generated via artificial intelligence (AI), compared to ** percent of those aged 18 to 29 years. Overall, the majority of U.S. citizens across all age groups expressed worries about AI-generated deep fakes.

Replication data for Political Deepfakes are as Credible as Other Fake Media and (Sometimes) Real Media. Primary data analysis and code work done by my co-authors Soubhik Barari and Christopher Lucas.

Recent technological advancements in Artificial Intelligence make it easy to create deepfakes, hyper-realistic videos in which images and video clips are processed to create fake videos that appear authentic. Many of them are based on swapping faces without the consent of the person whose appearance and voice are used. As emotions are inherent in human communication, studying how deepfakes transfer emotional expressions from original to fakes is relevant. In this work, we conduct an in-depth study on facial emotional expression in deepfakes using a well-known face swap-based deepfake database. First, we extracted the photograms from their videos. Then, we analyzed the emotional expression in both the original and the faked versions of the video recordings for all performers in the database. Results show that emotional expressions are not adequately transferred between original recordings and the deepfakes created from them. The high variability in emotions and performers detected between original and fake recordings indicates that performer emotion expressiveness should be considered for better deepfake generation or for detecting them.

Overview

With the rapid advancement of deep learning and generative adversarial networks (GANs), deepfake technology has seen significant improvements, making it increasingly challenging to distinguish real from fake images. To aid in the development of more robust deepfake detection models, we introduce a comprehensive dataset that combines and augments two widely used datasets, ensuring diversity, balance, and generalizability.

Dataset Composition

This dataset is built by merging and augmenting two existing datasets, each containing face images from distinct sources. By integrating these datasets, we ensure that the model trained on this data can generalize well across different domains.

1. 140k Real and Fake Faces Dataset

• Contains 140,000 images (70,000 real and 70,000 fake).
• Real images are sourced from Nvidia’s Flickr dataset.
• Fake images are generated using StyleGAN and sourced from the DeepFake Detection Challenge Discussion on Kaggle.

2. Fake-Vs-Real-Faces (Hard) Dataset

• Comprises 1,288 images (700 real, 588 fake).
• Fake images are sourced from thispersondoesnotexist.com, utilizing StyleGAN2.
• Real images are collected from Unsplash using its API, with faces extracted using OpenCV.

Augmentation Techniques

To further enhance the dataset and improve model robustness, we applied several augmentations: - Rotation - Shifting - Brightness Modification - Zoom - Cropping - Flipping

Through these augmentations, we generated an additional 6,445 images, leading to a final dataset of 12,890 images, where: - 5,890 are real - 7,000 are fake

Dataset Balance & Generalizability

Despite the vast number of images from the first dataset, we ensured a balance between the two datasets to prevent bias. The dataset is structured to equally incorporate images from both sources, fostering generalizability and preventing model overfitting to a specific type of deepfake.

Usage & Applications

This dataset is ideal for: - Training and evaluating deepfake detection models. - Studying the impact of GAN-generated images on facial recognition systems. - Exploring generalization techniques for fake image detection across multiple sources.

Conclusion

By carefully curating and augmenting this dataset, we provide a rich and diverse resource for researchers and practitioners working in deepfake detection. This dataset enables the development of more effective models capable of distinguishing real faces from synthetic ones across various sources and deepfake generation techniques.

Deepfake Fraud Loss Insurance Market Outlook

According to our latest research, the global Deepfake Fraud Loss Insurance market size reached USD 1.32 billion in 2024, reflecting the sector’s rapid expansion fueled by rising digital threats. The market is expected to maintain robust momentum, registering a CAGR of 32.1% from 2025 to 2033. By 2033, the global market size is forecasted to reach USD 17.87 billion, driven by escalating instances of deepfake-enabled fraud, heightened corporate risk awareness, and the increasing sophistication of synthetic media attacks. This growth trajectory is underpinned by growing demand for specialized insurance products that can address the evolving risks posed by deepfakes across multiple industries and user segments.

A primary growth driver for the deepfake fraud loss insurance market is the exponential increase in deepfake-related incidents targeting both individuals and organizations. As artificial intelligence and machine learning technologies become more advanced, the creation and dissemination of convincing synthetic media have surged. Financial institutions, corporations, and even government agencies are witnessing a rise in fraud attempts that leverage deepfake audio, video, and images to manipulate transactions, impersonate executives, or compromise sensitive data. This surge in sophisticated cybercrime has prompted organizations to seek dedicated insurance coverage to mitigate potential losses, fueling demand for deepfake fraud loss insurance products globally.

Another significant factor contributing to the market’s rapid growth is the evolving regulatory landscape and heightened focus on cybersecurity resilience. Governments and regulatory bodies across major economies are introducing stricter compliance requirements for digital identity verification, fraud prevention, and incident response. These regulations are compelling enterprises, especially those in high-risk sectors such as finance, healthcare, and e-commerce, to invest in comprehensive insurance solutions that cover emerging threats like deepfakes. The growing awareness of reputational, operational, and financial risks associated with deepfake attacks is driving organizations to adopt insurance policies as a critical component of their risk management strategies, thereby expanding the market’s reach.

Technological advancements in insurance distribution and claims management are also propelling the deepfake fraud loss insurance market forward. The proliferation of online platforms, AI-powered underwriting tools, and digital claims processing systems has simplified access to specialized insurance products. Insurers are leveraging data analytics and machine learning to assess risk profiles more accurately and offer tailored policies to diverse customer segments, including SMEs, large enterprises, and individuals. This digital transformation is enhancing customer experience, reducing administrative overheads, and enabling insurers to respond swiftly to the dynamic threat landscape, further accelerating market growth.

Regionally, North America dominated the deepfake fraud loss insurance market in 2024, accounting for over 42% of the global revenue, followed by Europe and Asia Pacific. The presence of a highly digitized economy, early adoption of advanced cybersecurity measures, and a robust regulatory environment have positioned North America as the leading market. Europe is witnessing accelerated growth due to stringent data protection laws and increasing investments in digital identity protection. Meanwhile, Asia Pacific is emerging as a lucrative market, driven by rapid digital transformation, increasing cybercrime rates, and growing awareness among enterprises and individuals regarding the risks posed by deepfakes. Latin America and the Middle East & Africa, while still nascent, are expected to register above-average growth rates as digital ecosystems mature and demand for fraud protection rises.

Coverage Type Analysis

The deepfake fraud loss insurance market is segmented by coverage type into individual, corporate, and government policies. The individual coverage segment is witnessing heightened interest as deepfake attacks targeting high-profile individuals, celebrities, and executives become more frequent. Individuals are increasingly vulnerable to identity theft, reputational damage, and financial loss due to manipulated audio and video content. Insurers are responding by