Deep fakes and big data: the next level of cyber warfare

Most people are unaware of the personal security risks confronting them when they make use of 21st century technology. Threats are embedded in the facial recognition technology that opens your iPhone, Chatbots on your online shopping sites, and unmanned aerial vehicles (UAV) deployed in military operations. These technologies form a juggernaut of sophisticated devices storing your most personal details and posing significant security threats. Every day, huge amounts of personal data are being vacuumed up by social media and communication apps alone. Platforms that do not provide end-to-end encryption, such as Twitter, Facebook messenger, and Snapchat, among the most popular social media apps, are particularly vulnerable to privacy breaches affecting consumers. Users of Chinese social media apps like WeChat and Tik Tok are at even greater risk as demonstrated by recent revelations of malpractice by the two providers. WeChat was exposed for secretly scanning users’ stored photos, without authorization. Tik Tok introduced a change in its U.S. privacy policy earlier, allowing the app to collect users’ biometric identifiers and biometric information such as faceprints and voiceprints. The two applications have active users numbering 1.2 billion and 700 million, respectively, according to 2020 statistics. Imagine the wealth of data amassed by the two apps, making unwitting users potential victims of mass surveillance, extortion, defamation, and disinformation. Indeed, the more collected data is available to feed AI algorithms, the more efficient and pernicious AI becomes in its capability to manipulate or abuse personal and biometric data. This creates a vicious cycle that continually advances the danger of data being deliberately used for criminal purposes. Deep fake damage doesn't end at defamation Deep fakes are constructed by machine learning and supported by a specific deep learning application known as the Generative Adversarial Networks (GAN), where two self-supervised algorithms can “learn” from each other. After thousands of training cycles, GAN’s algorithms have become skillful at producing synthetic images that are hyper realistic deep fakes, making them exceptionally difficult to detect with the human eye. Deployed for the purpose of spreading disinformation, deep fakes raise potentially disastrous consequences for leaders of democratic governments, as well as to international security. Domestically, deep fakes are capable of assisting cyber-criminals to commit crimes of identity theft, fraud, blackmail, and other malfeasance. All that is required is gaining access to the targets’ personal information and private data by deceiving the biometric security in smart technologies. For example, major banks like CIBC and RBC use voice verification security (voice ID) for their authentication services. Despite the banks’ repeated reassurances that their authentication systems are secure, the BBC conducted tests that defeated the voice recognition security system of HSBC by deception, in 2017. The test revealed that biometric security can be bypassed by sophisticated simulation, using deep fake technology. With sufficient data to feed the AI algorithm and the knowledge of how to apply it, a malefactor could impersonate literally anyone, doing anything in a fabricated video. Consider deep fake productions of former U.S. president Barack Obama’s profanity-laden outburst against Donald Trump or instances of celebrities performing pornographic acts. Perceptibly, when deep fakes are applied to spread disinformation, the effect may not end at defaming a public official and damaging his public image. Fair elections may be imperiled with resultant damage to the polars of democratic society. Given the relatively low cost, simple access, and authentic appearance of deep fake misrepresentation, the world is seeing more frequent and intensified use of the technology. State-sponsored disinformation runs at large through covert channels to stirring chaos among political and economic rivals around the world. It foments polarization, tearing the fabric of social and institutional orders. Additionally, when deep fakes are tailored to spark radical actions of certain target groups—for instance, diaspora communities, or even far-right groups and extremists—such consequences would not be limited to the former frequently becoming vilified and victimized, but also long civil unrest that fundamentally undermines the state’s authority and its liberal democracy. Advances in AI affect the whole spectrum of security, at the personal and international levels. The situation likely will get worse in the foreseeable future. The remedy must be to strike a balance between technological advances and how it may be used, without compromising security and freedom of expression. That, however, is the most challenging question of all. Sze-Fung Lee is a research assistant at McGill University. She holds a master’s degree in International Security from University of Warwick. Her research interests are in security policy, hybrid warfare, nuclear proliferation, and the politics of Hong Kong. Benjamin Fung is a professor and Canada Research Chair in Data Mining for Cybersecurity at McGill University. The Hill Times

Deep fakes and big data: the next level of cyber warfare

When deep fakes are applied to spread disinformation, the effect may not end at defaming a public official and damaging their public image. Fair elections may be imperiled with resultant damage to the polars of democratic society.

Health misinformation is hurting patients and undermining trust. Let’s do something about it.

News

Get free news updates

Opinion

Submit a guest opinion article

resources

Follow us

Advertise

subscribe

get help

Get free news updates

Submit a guest opinion article

Deep fakes and big data: the next level of cyber warfare

When deep fakes are applied to spread disinformation, the effect may not end at defaming a public official and damaging their public image. Fair elections may be imperiled with resultant damage to the polars of democratic society.

Become a Political Insider