In the modern age of digitalization, the world is always more than eager to welcome new technologies that offer people unprecedented advantages that make modern life a tad bit easier.
Over the course of recent years, however, as more and more enterprises ride the wave of digitalization and continue to integrate technologies such as cloud computing into their digital infrastructure- the use of Artificial Intelligence (AI) and Machine Learning (ML) has risen in the cybersecurity world as a staple in providing security to enterprises in an increasingly complex threat landscape.
However, unfortunate as it may be, the AI technology has often been exploited against enterprises, with statistics depicting a bleak picture, with more and more cyber-criminals turning to AI for launching increasingly sophisticated cyberattacks. One such dark side of AI definitely reveals itself in “deepfakes.”
If you’ve been following the slightest bit of cybersecurity news, chances are you’re familiar with the term “deepfakes.” However, if you’ve been spending the better half of the last couple of years in hibernation, the deepfake technology is basically every reason why some people are critical of rapid developments in the technology field.
Simply put, ‘deepfakes’ refers to the misuse of AI technology, that allows individuals to create highly realistic fake, or altered videos and audio files, which gives rise to cybercriminals and internet trolls equipped with the ability to mimic whoever they wish.
Keeping in mind the ever-evolving nature of the cybersecurity threat landscape today, the threat that the deepfakes technology pose is extremely dire. If the deepfake technology falls into the wrong hands (which it already has, considering the prevalence of deepfake ‘tutorials’ online), the tech could single-handedly topple over governments, and turn the world into an Orwellian nightmare.
Despite seeing snippets of the damage wreaked by the deepfake technology, most enterprises are still failing to realize the full scope of the threat posed by the deepfake technology, and just like the governmental agencies tasked with cybersecurity- are ill-equipped to detect and combat the deepfake threat.
Having said that, in an attempt to educate our readers about the full scope of the threat posed by deepfakes, we’ve compiled an article that delves deep into the damage deepfakes could potentially cause, and how a deepfake could be behind the next ‘big’ breach.
Are deepfakes really ‘that dangerous?’
Up till this point, the most widely circulated deepfake videos have mostly all been created with the intent to make audiences laugh- a fabricated video of actor Bill Hader morphing into Tom Cruise and Seth Rogen, an altered video that shows House Speaker Nancy Pelosi slurring her words, along with a hilarious video of actress Jennifer Lawrence giving a speech with her face swapped with Steve Buscemi’s.
In retrospect, however, even the mere existence of such videos begs the question- in the Information Age, can the authenticity of anything on the internet be trusted? Although none of the videos we’ve mentioned above were created with malicious intent, they demonstrate a far more disturbing notion and shine a candle on how easy it is to formulate and propagate misinformation on the web today.
When we take into consideration the upcoming U.S Presidential Elections of 2020, the concerns raised by several government officials seem to become more and more legitimate, especially as far as the danger posed by the deepfake tech is concerned.
Furthermore, the magnitude of the threat posed by deepfake audio and video files can be felt by the fact that although they seem to be targeting prominent politicians and celebrities, recent advancements in the deepfake technology could mean that anybody could be a potential target.
What’s even more alarming is the existence of revenge porn, which utilizes the deepfake technology for the malicious purpose of superimposing images of a targeted woman to create fake sexual explicit videos- which are then passed around to inflict lasting damage on the woman’s reputation.
Unlike the deepfakes of celebrities or the altered videos of President Obama being circulated on the web, even the mere notion of revenge pornographic deepfakes bears witness to the fact that anybody that has a picture or video of themselves somewhere on the internet, faces a distinct threat of being deepfaked.
In the Age of Information, as more and more enterprises ride the digitalization wave and rely on flawed (from a security perspective) technologies such as AI that many people regarded as a blessing or a curse and cloud computing, the attack surface area continues to expand exponentially. The dangers posed by the deepfake tech were also highlighted by a Republican senator, Marc Rubio who compared the deepfake tech to nuclear weaponry, stating that the technology could be used to “throw our country into tremendous crisis internally and weaken us deeply.”
How can deepfakes contribute to the next ‘big’ data breach?
Before we can get to factors that enable deepfakes to propagate the spread of data breaches and crime, we’d like to clarify what we mean by the phrase ‘big’ data breach first. For the purpose of explaining the grand magnitude of damage that deepfakes can cause, we’ve considered even the smallest of breaches to be significant as well.
With that out of the way, contrary to popular belief, deepfaking audio files can actually wreak more havoc than deepfakes rooted in the video, since victims are more likely to believe in a fabricated audio file, rather than a video.
The extent that people are willing to suspend their disbelief while receiving a fabricated phone call is reflected in how profitable the shady business of voice fraud is- with statistics reporting that the phone-based identity theft costs U.S consumers about $12 billion each year!
As seen in the real-life example where cybercriminals mimicked a CEO’s voice to demand an urgent ‘cash-transfer,’ cybercriminals and hackers can make a fortune, simply by fabricating a phone call, or a voice message which is a type of social engineering technique known as vishing. To prevent this, you can use a VPN of a reliable company. A VPN masks your real identity and avoid hackers from interfering with your communications.
With that being said, however, that although the manipulated audio files might sound the same to the average human listener, AI-centric cybersecurity programs can be used to detect the slight variations in pitch and speed, and successfully differentiate between real and synthetic audios.
So, where does the fight against the deepfake technology go from here?
Despite everything we’ve written up to this point sounding eerily similar to the plot of a sci-fi movie, there is still a lot of positive things to look out for, especially as far as the fight against deepfakes is concerned.
For starters, as more and more people are starting to educate themselves about the deepfake tech, and the implications the technology has on misinformation. At the end of our article, we’d like to remind our readers to check the authenticity of everything they see, hear or read online. In the world of deepfakes, being a responsible user of technology makes it a moral obligation upon users to determine the legitimacy of every bit of information thrown at them, instead of blindly believing everything!