Deepfakes are a fascinating (yet dangerous) phenomenon in our ever-changing digital world. Many of those responsible for spreading deceptive fake news have upped their game. Their work allows you to wake up to a video of you saying or doing something that never happened. Deepfakes can be detrimental to a person or business’ online reputation.
What are deepfakes?
Deepfakes are essentially audiovisual forgeries that can make any face appear to say or do anything. When deceptive content becomes indistinguishable from reality, it forces us to go against the law and our natures. Deepfakes differ from regular fake news in that they have biological roots. Humans rely on sight to interpret reality, so we are more likely to be fooled when we see something fake. In short, humans are wired to believe what they see.
Hollywood pioneered the technology that enables these forgeries. Years later, a Reddit user built on the same technological marvels that made Brad Pitt an old man in “The Curious Case of Benjamin Button” and Sam Worthington a Na’vi alien in the blockbuster “Avatar.” This user-shared code could be used to seamlessly superimpose a face over a body in a video.
Ironically, the first high-profile deepfake victims were Hollywood stars. A sophisticated forgery used famous actresses’ faces to distribute deepfake content to porn sites. Recently, a doctored video purporting to show Nancy Pelosi speaking during an interview while drunk was used to target her. We were all warned by the video’s subtlety, technical prowess, and intentional damage to her image.
The fear?
Deepfakes can cause untold harm. Long term, it may be possible to dissuade many from believing these deceptive videos, but the damage can be severe when used strategically. A day before a vote, a fake video can sway a tight election. A deepfake can cause a company’s stock to plummet suddenly and strategically. In reality, viral online content outpaces our response time.
The most vulnerable companies have prominent leaders who are linked to their reputations. It could go far beyond any damage control efforts, leaving a tenacious stain on both the individual and the company as a whole. Elon Musk recently announced his stock repurchase via Twitter, causing havoc in the market and affecting thousands of investors globally. Tesla and Musk were fined $20 million for this blunder. How would a crisis unfold if triggered by a fake video posted online?
Because deepfakes are unpredictable, Facebook and Microsoft have launched a joint program with the Partnership on Artificial Intelligence to identify and combat them. During the US election campaign, these tech titans began developing countermeasures to public pressure. There is also an ethical dimension to this issue: Do we share responsibility for controlling what content is published or shared?
Deepfakes will continue to threaten organizations until sufficient countermeasures are found. We must all be prepared to act quickly. In a recent interview with CNBN, Hao Li, a computer science professor at USC, predicted that in less than a year, anyone would be able to create “genuine” videos and images with no way to detect deepfakes. The Chinese app Zao is one example.
Most companies’ first reaction is to fight fire with fire. Companies like Symantec have investigated using artificial intelligence to detect deepfake videos by reversing engineering their creation. New startups like ZeroFOX, Truepic, and Proofmade focus on monitoring, verification, and social media protection to ensure accurate content. This rapidly growing industry now costs millions of dollars per year for high-quality services.
The issue with deepfakes is that they keep evolving, outpacing the solution. So, what are we to do?
Transparency and reputation are key.
Contrary to popular belief, communication and crisis management teams used to fight lies by controlling the narrative.
But we can no longer rely on the Aristotelian sender-message-receiver model. This idea served as a guideline for accurately conveying the sender’s message to the receiver, but it is no longer applicable. Aristotle’s model has finally crumbled. Now we must consider how we will seize control of the narrative when the receiver cannot tell the sender from a liar.
The corporate world today is ill-prepared (at best) and completely unprepared (at worst). Even the most innovative committees and political advisors have barely touched this topic. Anyone relying solely on pre-deepfake crisis management systems will be in for a rough ride.
This new paradigm pushes our already VUCA (short for volatility, uncertainty, complexity, and ambiguity)world to the limit. This forces us, strategists, to create new tools and methods to protect companies, public figures, and organizations. A modern action plan must include the following elements:
Know Your Enemy. How does a fake news story go viral? What policies do Facebook, Twitter, and YouTube have to prevent them? Can one ask a platform to remove a video? What does such a request mean legally? Who are the most receptive audiences?
Know Your Weak Points. What personal or professional areas could a deepfake news creator target? Where would these misinformation campaigns come from? Is your reputation trustworthy enough to deny any misrepresentations? A thorough self-diagnosis is required to anticipate and respond quickly.
Assemble a Crisis Team. Forming a crisis management team is an essential first step. A company’s communications director must regularly communicate with its legal advisors, director of Operations, spokespeople, and Social Media team. Policymakers will be able to develop solutions based on clear definitions and strategic guidelines.
Set up Incident Channels In damage control. It can make a difference whether the information is spread through your channels or by an influential journalist. The most efficient and wide-reaching channels to use in case of negative virality.
Locate Contacts. It could be an official social media account or editors of a well-known media source. These parties can all help mitigate the effects of a deepfake. You must establish reliable contacts so major media outlets can quickly debunk fake stories and stop their spread.
Find digital experts. A team that understands technology and the digital world will be the most effective in detecting and exposing deepfakes.
Communicate clearly. Providing brand communication templates may seem routine, but in a crisis, every second counts. Prepare communication materials that can be quickly adapted and published as part of the crisis management process.
Grow a Digital Identity. Not unlike football or soccer, a great offense is produced by having a great defense. The first line of defense for any company is a robust online reputation for both the company and its leaders. This can help speed up recovery and minimize damage. Why? Uncertainty makes it easier to debunk falsehoods when a respected CEO is transparent. Also, citizens, journalists, and opinion leaders will believe the CEO when they state the truth.
Companies today must devise strategies to combat this insidious form of misinformation. The first step is to conduct an internal audit to fully understand the company’s capabilities and reach, followed by crisis management training for all spokespeople. As always, we believe anticipation is critical.
In Closing
We must fight misinformation and the sea of lies, but we must also prepare for the worst. Online reputation management will become more difficult in the future if we cannot harness the world of deepfakes.