Share this...

By Trisha Prabhu

“Hi Trish, I’ve been hearing all about deepfakes in the news these days. What are deepfakes and why are people so worried about them?”

Hi there, and welcome back to Ask Trish! I hope you’re having a wonderful start to the week (and, depending on where you are in the world, possibly beginning to enjoy some gorgeous fall weather – my favorite season!).

Thank you so much to this week’s question-er for the fantastic question. Believe it or not, to date, we have not actually covered deepfakes on Ask Trish…between that and all the buzz around them (and, as we’ll discuss, their potential implications for society), I think it’s incredibly important that we do a brief deep dive into deepfakes. Hopefully, in doing so, I can offer y’all some clarity on what is ultimately a pretty complex topic. To the questioner – I promise you’re not alone in wondering what deepfakes are and why they’ve got so many folks scratching their heads/pulling their hair out. While this post won’t offer a comprehensive look at deepfakes (as always, please use these posts as a jumping off point with which to learn more about the topic at hand!), I will offer you a clear, simple look at 1) what deepfakes are, 2) why we’ve seen them in the news lately, and 3) why folks are worried about them. I’ll leave you to ultimately assess the risk they pose!

Sound good? Let’s get into it!

Let’s start with the basics. What are deepfakes? Well, as you might guess, deepfakes refer to visual and audio content – often content that depicts a person/persons’ likeness or face (though the content can also depict other things, like events!) – that is, in fact, digitally manipulated. That explains the “fake” part of deepfake…but what about the “deep” part? Well, the “deep” part refers to the fact that deepfakes are generated via a form of artificial intelligence called deep learning. Even if you’re only a little familiar with AI, you’re likely aware that – especially in recent years – it has become pretty darn good. The result is that many deepfakes can and do convincingly portray people and events that are totally false…and it can be hard for the average internet user to know that. #yikes Today, deepfakes can seemingly put words that someone never said into their mouth, make it seem like someone can do something, e.g., dance that they definitely can’t do, or suggest that events that never happened did happen.

Okay, so now we understand what deepfakes are. Why have they been in the news a ton in the last few weeks? Well, in the last few weeks, there have unfortunately been a slew of AI deepfakes targeting celebrities – and it’s led many celebrities to speak out against deepfakes. Gayle King, the popular host of CBS News, spoke out after her likeness was used in a video advertising a weight loss product; “please don’t be fooled by these AI videos!” she said. Actor Tom Hanks similarly spoke out after his likeness was used in a video advertising a dental plan. He too had never heard of the plan and wanted to warn his followers/fans that the whole thing was a dupe. (Frustrating and scary, am I right?) In many ways, though, King and Hanks’ statements – and the buzz around them – are just the latest in what’s been a lot of news coverage around deepfakes, especially as they’ve been used to make it seem like politicians said or did something that they did not, in fact, do. (You can imagine how that might get messy, especially during an election season…)

Given all of that, it’ll likely come as no surprise then, why folks are so worried about deepfakes. For one, people are concerned that they may promote disinformation, as in the case of the false advertisements featuring King and Hanks. Weight loss products are one thing, but what about deepfakes that falsely portray violence or clashes that ultimately stir emotions and lead to war? There’s also concern that deepfakes are disproportionately weaponized against women and amount to violence against women. In fact, of the 15,000 deepfake videos AI firm Deeptrace found on the internet in September 2019, 96% were pornographic, and 99% of those were content that mapped the faces of female celebrities onto porn actors. This, no doubt, should give us pause. Still others worry that deepfakes will threaten the work of creatives – actors like Tom Hanks himself. (Some believe that 50 years from now, all movies will be entirely AI-generated.) It raises the question: is that fair? Is that the world we want to live in? (Perhaps it is, but no doubt, we’ll need to have some conversations first…) Note that there are many, many other issues that folks cite too, but these are three that you’ll commonly hear. Folks disagree on to what extent these concerns pose a real risk to our society – that’s where things get tricky – but at a high level, this is why people are worried about deepfakes.

With that said…I hope that this was a helpful introduction to deepfakes. Undoubtedly, there’s a lot more to this topic, so I hope this post inspires you to do some internet research and learn more about this complex issue.

Now, as always, before I wrap up, I want to encourage y’all to share any Internet-related questions, thoughts, or perspectives (whether deepfake-related or not!) here. Your contribution just might be featured in an upcoming TikTok/blog post! Not only do you get some helpful advice, and an absolutely incredible Ask Trish video (am I right?), our entire Ask Trish community benefits, too. If you share a question, you might just inspire someone with a question of their own to do the same! So, please – take just 30 seconds (#nolie, that’s all the time it takes!), and fill out the form. Thanks in advance for contributing to our #AskTrish community.

Wishing you a great rest of the week,


Share this...