What is a deepfake?
A deepfake is a doctored video of a person that appears to be an original. Like this video of Barack Obama. This, of course, is not Obama, but a “deepfake” Obama. Deepfake videos can show us fake footage of people acting, looking, and sounding the same as if we were watching real footage of them. Think Photoshop, but for video.
Deepfakes use a machine-learning model that analyzes video of a person, until the algorithm can create completely new, fake videos of the person. It is like putting on a mask that sticks to your skin and lets you “become” the person you want to imitate.
The deepfake can completely change how we perceive reality and add a whole new layer of distrust in video communications.
Deepfakes were born in Hollywood, and made accessible to all thanks to the enterprising underbelly of the internet. In December 2018, it is believed deepfakes swayed a presidential election in Gabon. In 2019 we witnessed a seemingly inebriated Nancy Pelosi struggling to answer questions shared by millions and even President Trump himself. And recently, a fake video surfaced of Facebook CEO Mark Zuckerberg looking straight into the camera and claiming “control billions of people’s stolen data, all their secrets, their lives, their futures”.
The implications of deepfakes
It is not the first time our skepticism has been put to test. Doctored images have existed as far back as the 19th century. However, this test will be different. Deepfakes add a layer of never before seen reality to disinformation.
Let’s take a look at some of the possible implications of this new reality:
- Attempts to regulate deepfakes: Republican senator Marco Rubio warned that deepfakes would be used in “the next wave of attacks against America and western democracies.” These fears have prompted Republican senator Ben Sasse to introduce the US’s first bill to criminalize deepfakes.
- More stringent content review on social media: According to the latest Edelman study, people’s trust in social media is at only 41 percent globally. The rise of deepfakes will put even more pressure on social media platforms to regulate shared content. An unintended consequence is that social media platforms (which already has been accused of bias) could leverage deepfakes to impose even heavier content policing. In doing so, the platforms themselves will have greater control in determining acceptable and unacceptable content.
- Content verification badges: Social media and online outlets might add “deepfake” verification badges to ensure viewers the video is authentic. “Unverified” content may raise eyebrows and be ignored.
- The rise of deepfake verification companies: Faculty is a London-based startup specializing in AI and machine learning. They generate thousands of deepfakes to train systems to distinguish real content from forgeries. Amber Video, based in New York, is on a mission to clean up the internet from disinformation with software embedded in smartphone cameras to act as a verification watermark.
- Deepfakes will become easier to create: 28-year old James, founder of the derpfakes Youtube channel transposed actress Carrie Fisher’s 1977 face into the 2016 Rogue One movie in “the time it takes to watch an episode of The Simpsons.” The technologies behind the AI that makes deepfakes possible will become more sophisticated and the mechanisms to create the deepfakes will become easier to use, allowing a larger percentage of the population to create and distribute them.
The first line of defense against deepfakes is having a response team ready to identify the proliferation of the fake material so that immediate action can be taken
Implications for personal brand reputation
CEOs, politicians, and others in the public eye recognize the dangers of deepfakes. At the minimum, they must implement systems to continuously monitor their image online. However, more sophisticated organizations will develop a response team ready to identify the proliferation of the fake material so that immediate action can be taken. Leaders in the public eye might even become more cautious in appearing in public since the exposure can make it more likely their image will be tampered with.
We will possibly see all of the above implications arise to the surface within the next few years. We will witness the relevance of deepfakes on a large scale for the first time in the 2020 US elections. It is possible that authentic videos will be discredited as fake and vice-versa.
If we don’t know what is real today, a few years from now it will be even harder to answer the question. We will have to learn to think more for ourselves and put a critical eye on absolutely everything we see online or offline.
DON’T GET CAUGHT BY SURPRISE
We’ll send you our monthly newsletter, The Angry Mob, detailing stories of leaders on the wrong side of the public’s pitchfork.