ZeroFox Expands AI Capabilities to Detect Deepfakes
Expanded best-in-class artificial intelligence protects against rapidly evolving threats across all social media and digital platforms; releases open source toolkit, Deepstar, to aid security community in building deepfake defenses
LAS VEGAS, August, 8, 2019, ZeroFox, the leading provider of advanced artificial intelligence-powered digital risk protection, today announced the latest evolution of its artificial intelligence (AI) capabilities, with the release of new video analysis features. The initial use case for this AI-based analysis is the detection of deepfake videos–an emerging threat that can be leveraged in misinformation campaigns. In addition to deepfake detection, ZeroFox is donating Deepstar, a new open source toolkit to help research teams and the greater cybersecurity community to tackle the new threat posed by deepfakes and enhance the accuracy and scale at which these detection capabilities must operate.
As attackers adopt advanced image-based techniques to evade traditional text, link and file analysis-based security solutions, ZeroFox’s new ensemble of AI-enabled computer vision and video analysis features eliminate the manual, time-intensive process of analyzing millions of images and videos that pose significant threats to brands and businesses.
“The challenge of protecting organizations from digital risks is becoming more difficult every day. Attackers are now composing and manipulating videos and leveraging images in ways that evade the detection capabilities of legacy solutions,” said James C. Foster, CEO of ZeroFox. “Over the past several years, we’ve been committed to harnessing the power of AI and applying it to the detection and remediation of modern digital risks. From OCR to image identification, and now advanced video analysis, we are able to efficiently and effectively identify and remediate critical risks at global scale.”
Social media and other digital platforms have rapidly increased the creation and dissemination of image and video-based content. Bad actors have taken advantage of these mediums, posting threatening images prior to physical attacks, posting images of cash and credit cards to conduct scams and make money off legitimate financial institutions. What’s more, the rise of deepfakes promises to further weaponize misinformation through altered videos. With the introduction of these advanced AI capabilities, ZeroFox is equipping teams with the ability to identify new threats, recognize brand infringement or impersonations, and remediate malicious or threatening content.
Expansion into AI-Powered Video Analysis
ZeroFox is a pioneer in leveraging AI to provide customers with advanced solutions to modern threats. In the past year, the team has introduced new AI capabilities, including technologies such as text analysis (sentiment analysis, NLP) and image analysis (object detection and image comparison). With the introduction of the new video analysis capabilities, ZeroFox is providing solutions that can analyze content and identify risks at scale across mediums. Benefits offered by the new capabilities include:
- Analyze threats in images and videos
- Identify and remove threatening images or leaked credit card information
- Detect deepfakes
Combating and Understanding Deepfakes
At Black Hat 2019 in Las Vegas, Nevada, ZeroFox’s CTO, Mike Price and Principal Research Engineer, Matt Price are discussing how deepfakes can be leveraged for offensive and defensive purposes in their session, “Playing Offense and Defense with Deepfakes.” The team is announcing the release of a new open source deepfake toolkit, called Deepstar, significantly reducing the time and toil required to produce deepfake detection capabilities. Deepstar includes code for automating the creation of deepfake datasets, testing, and enhancement of detection algorithms, along with a curated library of deepfake and real videos from YouTube. The toolkit incorporates a plug-in framework, enabling researchers to easily test or re-train and compare the performance of different classifiers. This is an important toolkit that enables researchers and the greater cybersecurity community to build and improve defensive capabilities.
“Today, anybody can download software to produce a deepfake video. As a result, deepfakes are cheap and easy to create and we expect bad actors will take advantage of these economics for nefarious activities,” said Mike Price, CTO of ZeroFox. “With the release of Deepstar, researchers and defenders will have an additional tool in their toolkit to assist in streamlining the process of deepfake detection research. With the likely abuse of deepfakes as part of an effort to misinform the public, we felt it was important to contribute our toolkit back to the community that has already done some great work, and to help defenders improve their ability to prepare for future challenges in this area.”
To learn more about Deepstar, please visit our blog: zerofox.wpenginepowered.com/blog/detecting-defending-against-deepfakes