Preparing for Deepfakes
Deepfakes, the fake news stories, photos, audio clips or videos of people saying things they didn’t, or depicting events that didn’t happen, are one of the many new realities of the age of artificial intelligence. As with any technology, AI has two side: the good and beneficial and the bad and nefarious. They’re kinda like the “Good Witch” and the “Bad Witch” in The Wizard of Oz – only with much more serious implications.
The Internet is full of discussions about deepfakes. Due to our current presidential contest, most of them are focused on the impact they can have on our political process. Canada’s CBC News recently released a good examination of this issue. In addition, an older BBC piece also showed how real deepfakes can be. And that was 6 years ago! Clearly, AI has advanced at a remarkable pace since then.
We should accept the fact that deepfakes are going to be part of the political juggernaut we are now beginning to experience in the run-up to America’s General election in November. If the soon-to-be-created deepfakes aren’t created by political opponents of particular candidates, then we will be exposed to deepfakes created by unfriendly foreign governments in an attempt to disrupt our society and way of life. Exploiting our already divisive environment is clearly part of the strategy focused on weakening America.
Just as a wedge can be driven between us through the use of deepfakes aimed at our political process, so too can they be used to fracture the relationship between the police and the people they protect. I’m sure there are already instances in which deepfakes depict or report fake police misconduct. I’m absolutely convinced that there will be more in the near future, and, that generally, policing is unprepared for them. Anticipatory police leaders will already be educating themselves and their staff on deepfakes. They will also be developing strategies for dealing with deepfakes aimed at their agencies. Part of this planning is developing a public information strategy for helping the public understand the deepfake phenomena. Before they occur. Helping the public understand what deepfakes are, the devastating potential impact they can have on the trust and confidence people have in the police, and how communities can dodge the malicious intent of deepfakes, is the best preemptive approach policing can take.
We have created an infographic to help policing begin educating their communities about deepfakes. You can access it here. It’s predicated on the belief that policing will be the victim of this technology. Further, that encouraging the public to think critically about disturbing images they may see, or writings they may read, and asking themselves if they square with what they believed before they were exposed to the deepfakes, may give policing leaders time to publicly react to them before their impact is intractable. I’ve often said that the time to “invest in the bank of community trust” is before you need to make a withdrawal. This is definitely true about inoculating the public from the impact of deepfakes.
This is such an important issue that we will revisit it many times in the coming months. If you have other anti-deepfake strategies you’d like to share, please reach out via our Contact Us page or by clicking here.
About the author: Jim Bueermann is the president of the Future Policing Institute and the Center on Policing and Artificial Intelligence.
