The Privacy Breach Nobody's Talking About – Google Gemini

Is Google's Gemini AI reading our emails and private chats? So this is what this lawsuit is about. A tech journalist said that Gemini revealed his first crush in elementary school and revealed one of his personal character flaws. Welcome to another episode of Silicon Valley Front Row. You're definitely getting the front row with us. And David Lam, Steve Vispas, we're going to talk to you today about, we're just going to jump into it, David. No more haircut stories. We already talked about the last time. Dang it. Yeah, let's just jump in. So this is big. This is serious, right? As much as we, I mean, who doesn't use Google, right? Use it every day. Yeah, so there's a, we'll use it right here. So Google is being sued for allegedly using Gemini AI, which anytime you search in Google, it's by default, it's using Gemini, their own AI tool, for basically for privacy issues, for invading people's privacy. So what's going on? This is a... Yeah, so the big question is, is Google's Gemini AI reading our emails and private chats? So this is what this lawsuit is about. And a tech journalist said that Gemini revealed his first crush in elementary school and revealed one of his own personal character flaws. And this was alarming to the journalist because he said he was just surprised by this when he was chatting with Gemini. So here's details about that. So until around October 10th, Gemini AI assistant required people to opt out or to opt into its feature. But after October 10th, this feature was allegedly secretly turned on by Google for all its users, Gmail, chat, and Meet accounts by default. So this enables AI to track the private data of people for those platforms and quote, without the user's knowledge or consent. And that's according to the complaint that was filed November 11th. So I'm not gonna ask you which media the journalist is with. Maybe it doesn't make a difference. But to have that kind of detail, like who he had the crush on in elementary school and stuff like that, that's pretty crazy. Because I mean, if that's being revealed, what else is gonna be revealed? I wanna play a quote or a soundbite from Joseph Gordon-Levitt. He's a male American actor, I think known for Third Drive from the Sun, Looper, 500 Days of Summer. Yeah, I like him as an actor, but he basically posted this video addressing his statements on AI regulation. Let's take a look. AI chatbots being allowed to prey on kids is not a matter of national defense. And yet right now, a bunch of big AI companies are trying to convince our government that it is. I've just been corresponding with a bunch of people in DC who work on this. I'll tell you what's happening. It's got me kind of angry. These companies are very sneaky. Here's what they're doing. Every year, Congress passes this thing called the National Defense Authorization Act. I'm just learning. It's basically how they allocate the budget for the military, et cetera. And it passes every year. It's about to pass this year. But at the last minute, after the representatives and the senators have all already voted for it, they're trying to sneak in this new provision that says states are not allowed to make laws regulating AI. All right, so Joseph posted that on November 18th. And it's concerning regulation laws around AI. Yeah, and we've seen celebrities and people just make statements publicly about their stance on AI. So a little bit about this class action lawsuit that is in California. It alleges that Google is violating the California Invasion of Privacy Act, which is a 1967 law that prohibits surreptitious wiretapping and recording of confidential communications without the consent of all parties involved. It's very interesting. So when it came to that journalist, his name is Thomas Tell. He's the plaintiff. He suspect that his private information was just been exposed by Gemini. Wow, that is really interesting. I have to say it's, almost every day we are surprised by some of the developments that happen with AI. And this is even something that's kind of run by a company. We're not even talking about the bad actors that could be using AI to do stuff like this. This is Google doing it. I think there's more to the case, right? Because I think it's good to understand more of the details, like how it actually works. We've been trying ourselves to figure this out. It hasn't been very easy. Can you explain like how this actually works? Yeah, so apparently the AI feature is found in the Google Workspace smart features in the settings. So once it turns on, so I tried to find it, I couldn't find it, but. Yeah, you try to disable it, right? Yeah, I'm not that techie, so anyway. I grew up with like laptops. I grew up when I was in third grade, had a laptop, you know, use that. But even then it was hard for me to find the setting too. But when I did, it said that my administrator of my email account has to be the one to change the setting. So even then I can't use it, that's my work email. So it was like, yeah, a lot of, like you mentioned off camera, like lots of hoops to get through. You have to jump through hoops to get it done and it's unbelievable. Yeah, so that's how it works. And apparently if it's turned on, which I guess it is by default. By default is on, that's the big thing, that by default is on. And it can scan, read and analyze every email and email attachment, message and conversation on those services or apps. Do we have a sense of how that information is being used? In the sense, can someone from Google access it and use it in some way or it just sits there? Well, what we know is that it's used by, for the service. So by the AI and it could be used for multiple things. To feed you ads or whatever. Possibly. That's like a privacy issue right there. So anyways, maybe one of these days we'll do a demo to show people how to get to the settings. Because we tried it just earlier and it was pretty difficult for us. So we just wanna get this episode out for you. But maybe we'll do another follow up on this. And we wanna show you here, this is, we wanna show you here some of the fine prints that Google has. And this may be one of the questions for our audience. How many of you go through the fine prints? Which are pages long in small font to explain some detail about your rights as a user. But anyways, in the fine prints they are saying that they are committed to privacy and all that. Which obviously is not the case. There's a quote from Google actually. So what they said is, we do not use your workspace data to train or improve the underlying generative AI in large language models that power Gemini, Search and other systems outside of workspace without permission. But it seems that if it's turned on by default. They can take that as permission. Apparently. Pretty amazing stuff. And a lot of loopholes, a lot of language that depending on how you use it. But the biggest thing I get from this is how that it's by default on. And again, you have to go through complicated process to turn it off. And think about this, you have what billions of users using Google. There might be some who will turn it off. But what's the percentage of that? My guess would be 0.01%. They will go through the effort to turn it off, right? So the majority will not. That's pretty amazing. Well, what are the questions for our audience on this one? So the questions for our audience, well, multiple questions. But the first one is, are you concerned that AI could be reading your chats or emails? And the second question is, are you worried about the privacy that you may be losing when you use Google Gemini or Google by itself? Do you have any concern about that? Of course, you may answer back saying that, what choice do I have? If that's your answer, please put it down. Exactly. And we're going to read some comments we got from our viewers. This one is by someone, by the username of Bo Carlson Third. Thanks again for your comment. And this is in response to a short regarding the indie filmmaker story about Tilly Norwood. Yeah, that's the episode we did. So they said, the difference between AI slop and art is that art has your voice and what makes you special. People want to see your vision and not some generic bland mush. Most people would rather see something unpolished or outright bad, but have something that's uniquely you than watching a slick AI nothing burger. We appreciate that comment. Yes, originality and human interact, human elements are crucial in art. And we're glad they see that. Yeah. And this reminds me of back then when we had to do school projects, like make the finger painting turkey or make these small arts and craft projects as kids. It might not be perfect, but there's memories and experience from it. And our families cherish that. That's right. Well, again, thank you for your comments. Keep them going.