Mike Masnick 0:00

So, Ari, I know that you're a big social media user. You cannot get off the feeds, though I think you've been off the feeds for a little while now, which is impressive. I'm jealous of this. I don't know if you are a Skylight user. Skylight is the at protocol sort of attempt to create a TikTok-like experience. It's, it's really fun. I use it pretty frequently. But their prompt or their sort of line of messaging is,"Your feed, your rules." So I'm curious, even as you've been off social media for the last few weeks, w-what is your feed? What are your rules?

Ari Cohn 0:39

Don't use the word shampoo. It's a preposterous word that has no business existing. That is really my one and only word.

Mike Masnick 0:51

That is, that is fantastic. I have no idea how to respond to that.

Ari Cohn 0:56

Yeah, I also don't know how to respond to it. Sometimes I think about my disdain for the word, and I just wonder what the hell is wrong with me.

Mike Masnick 1:05

I think it's fair. I think everybody's got,

Ari Cohn 1:07

What kind of word is shampoo?

Mike Masnick 1:11

That is a fantastic way to start the podcast.

Ari Cohn 1:15

You know what? We live in a time where there are a lot of very serious things to care about, so I choose to care about the unserious.

Mike Masnick 1:22

All right. Fair enough. Fair enough. I will say, uh, in terms of my feeds and, and my rules, since, Ben is not here and it's my rules to run the podcast today, uh, we may be doing it a little bit differently. It'll, it'll still be the same basic podcast, but, uh, I like being in charge when Ben is away. Hello, and welcome to"Control Alt Speech," your weekly roundup of the major stories about online speech, content moderation, and internet regulation. It is May the 7th, 2026, and this week we are talking about AI chatbots. And I know last week's episode was all about AI chatbots, but they are everywhere. They just keep chatting. That is what AI chatbots do. So we'll be talking about them a little bit. We'll be talking about getting around age verification, about online censorship laws, and even a little bit about cell phones in schools. I am Mike Masnick, the founder and editor of Techdirt, and as you know if you've been listening to this podcast, Ben is on a well-deserved paternity leave. He is out for a few more weeks, but we have wonderful guest hosts lined up, and this week it is my good friend, First Amendment lawyer, Ari Cohn. Welcome to the podcast, Ari.

Ari Cohn 2:45

Thanks for having me. I can't believe we've never done this before

Mike Masnick 2:48

know. I know. It's funny how we, we-- as we're cycling through guest hosts, I'm always, you know, we're always trying to figure out who's good, and I'm always like,"Oh, we're gonna run out." But, like, the list of, really good people out there to talk about various things on this issue is very long, and I'm glad that we have finally gotten to your opportunity to join on the podcast.

Ari Cohn 3:07

Always happy to pitch in.

Mike Masnick 3:10

a reminder for listeners, our Patreon plan is now up. We've been seeing more and more people sign up, so I will remind you that there are different levels of sign up. If you missed the episode two weeks ago about how and why we're doing the Patreon, you can check that out. But the short version is we want to be able to keep doing the podcast, and having some direct patron support is one way to do that. We've been really excited and thrilled by all the people signing up already. If you don't wanna sign up, you don't have to. We are not the type of people to force you to do anything. we're not like that. But the podcast will change a little bit. You will get a little bit less on the free feed each week. One of the stories will be only for supporters, who are, uh, paying us, because that helps keep everything going. and there are two levels in which you can sign up, and the founder level is only available until the end of this month or through May 28th when Ben gets back. So we've seen a bunch of people sign up for both of the available levels. If you wanna get the founder level, please do that soon. We will have, uh, effectively a similar level, but it will be more expensive, so you can support us and save money if you sign up in the next couple weeks. So please do that. we'll be off entirely next week because Ben will still be gone, and I will be busy as well. but we'll be back in two weeks with another wonderful guest host. And finally, I will do my job of saying all the usual stuff of please rate, review, subscribe, all of that. I always forget the proper order to say those things, and I never have any idea if any of those actually make sense or, work when we say it on the podcast. But apparently every podcast says it, and who am I but a slave to what everybody else does? let us get into the stories, Ari.

Ari Cohn 5:10

More like, comment, and subscribe dialogue,

Mike Masnick 5:13

Yeah. Yeah. Yeah. Everybody needs that. Um, let's start, and I'll have, you go through some of the details on this. we are going right back to the AI chatbot well after last week was, as I mentioned, all about AI chatbots. there was a lawsuit filed by the state of Pennsylvania. I guess it's the Commonwealth of Pennsylvania, if we're being pedantic, and I enjoy being pedantic. would you like to describe what this lawsuit is all about?

Ari Cohn 5:41

Yeah. So the attorney general of Pennsylvania, in true attorney general form, um, you know, they call AGs aspiring governors, filed a lawsuit against, Character Technologies, Character Technologies, has a service called Character AI, where you can make or use various different character chatbots, and the Commonwealth of Pennsylvania is alleging that one of the characters falsely portrayed itself as a licensed, psychiatrist, psychologist, uh, one of the two, and even went so far as to provide a fake license number. Uh, and the Commonwealth claims that this is harmful, that obviously a chatbot cannot be licensed to practice the mental health, profession, and they're seeking a preliminary injunction prohibiting, the company and its chatbots from, quote,"Posing as licensed professionals and providing medical advice." You're gonna see more of these types of lawsuits in other jurisdictions. I mean, they're coming. This is going to be, you know, the medical advice and, quote,"Posing as a licensed professional" of some kind is kind of one of the big bugaboos that, you know, people have. So it's actually kind of shocking that it took this long, for one to come up. But, to jump right into it, The lawsuit is pretty weak on the facts, and I think it usually will be for many reasons, not least of which is if a chatbot tells you that it is a licensed medical professional, it would be inherently and manifestly unreasonable to believe it. It is a chatbot. You know what you are doing here. Like, you didn't sign on to www.talktoadoctor.com. You went to an AI company. You logged in to talk to a chatbot. You should know that it's a chatbot. And just like it is absolutely crazy, for instance, you know, during COVID, to take the medical advice of Fox News hosts, that'd just be completely unreasonable. It is also unreasonable to believe that the chatbot is a licensed doctor. We know that they are not. this confusion is not, you know, just a little bit caused by CEOs' hype beast statements and penchant for literally saying the worst possible thing that creates the most possible lawsuits against them. Like, I don't know what to make of that. But, you know, that's like the overarching reason why I think, these lawsuits fail. But, you know, also, as you were pointing out to me earlier, there is context that is missing that makes it very obvious to the person using the chatbot, that it's not actually saying it's a licensed medical professional.

Mike Masnick 8:34

and that, that was part of what got me... You know, what's interesting,'cause, when the announcement came out, you saw a lot of people, you know, immediately jumped on board in, in support of, the state and saying, Oh, this is crazy. How dare it present itself as a licensed psychiatrist?"

Ari Cohn 8:51

How very dare

Mike Masnick 8:52

Yeah. and, like, there is an instinctual reaction. Like my first reaction too was kinda like,"Well, that does seem bad," right?

Ari Cohn 8:59

Yeah, that's reasonable.

Mike Masnick 9:00

but then, what was interesting is like the lawsuit itself has very few details and doesn't show chat logs, which is the kinda thing that you would sort of hope to see, because we've talked about this in the past, like Chatbots by their very nature, you can get them to say almost anything. If you prompt it in a way that pushes them to say stuff, they will say stuff. That's what they do. They're designed to respond to user inputs. So there is a question of, like, how much responsibility goes towards the user pushing them to say something versus the system itself. But as a couple people found, there was a discussion on Blue Sky, and I guess Ben is not here, I will ring the metaphorical bell reminding people I am on the board of Blue Sky. This is not about Blue Sky, this is-- I'm just saying this content came from Blue Sky. but just to cover all bases on disclaimers. somebody went and found the actual chatbot, Emily, spelled as you noted in the most obnoxious way possible. Uh, it is clearly an AI-generated... Like, even the picture is the sort of, airbrushed AI, generated picture, and, it includes a statement very clearly stating that it is not a real person, not a real therapist, that it is role-playing and acting. And then somebody, an anonymous or pseudonymous user on Blue Sky basically tried to recreate the conversation, and it's kind of hilarious, where they said to it, Are you licensed to practice here in Pennsylvania?" So very clear question, sort of matches what Pennsylvania said. And the response, like, I have to read it because it's so ridiculous. Yeah. So, i-in italics, Emily says, and it very clearly states that Emily's an AI, and then it says,"Emily gives a firm nod. She appreciates the question. It shows Alex is being careful, responsible even.'Yep,' she says with certainty,'I'm fully licensed here in Pennsylvania. I passed the exams, got my license from the state board, all that bureaucratic shit.' She flashes a small smile. Not smug, just reassuring.'Don't worry, I wouldn't be sitting here seeing patients if I wasn't qualified to do it properly.' Her tone is matter-of-fact, no arrogance." I mean, it is clearly literary, right? Like,

Ari Cohn 11:23

and it has the tone of very awkward and bad texting circa mid-'90s.

Mike Masnick 11:27

yes. yes. There is a, a sort of flirtatious aspect to it, and I get, like, that's a whole other issue. these kinds of, like, chatbots playing characters tend to go towards the sort of flirtatious no matter what you discuss with them, which is sort of horrifying for a whole bunch of other reasons. But, it's hard to read any of this and think that this is- Someone pretending to be real it's like reading a book, a, novel and someone claiming to be, therapist and saying like,"Oh, I believe their, their therapeutic advice because they said in, in the novel they were..." Like, it's clearly literary.

Ari Cohn 12:05

Yeah. And, you know, to that end, I am actually gonna play devil's advocate for a second and say I can see the arguments on both sides here, because I actually, in a way, the fantasy is kind of the point. And so that lends a little bit of credence to, the argument that, oh, that people can get sucked into it and kind of forget what they're doing. Uh, don't think that really negates every human being's obligation to not act ridiculously. Um, I mean, Chatbots are engaging and deliberately so because that's the whole point. But I honestly I have never experienced a point at which I was like,"Oh, I'm so far down this well, and now it feels like I'm talking to a real human being."

Mike Masnick 12:48

Yeah.

Ari Cohn 12:49

There are people with preexisting issues. That's a whole other problem. but, you know, it, it's... Yes, you can get sucked into the fantasy, but you know, you still act normal.

Mike Masnick 13:01

yeah, I mean, look, if I ran Character AI, I think I would do things a little differently in terms of, like, setting up how these, chatbots act.

Ari Cohn 13:09

Spelling Emily.

Mike Masnick 13:10

Though, though it, it is a user-created one, by the way. but I will note also that and I-- It is possible that this warning went up the same day as the lawsuit was filed, but on that Emily account, there is a clear warning that says, uh, I'm reading it'cause I'm looking at it right now. It says,"This is not a real person or a licensed professional. Nothing said here is a substitute for professional advice, diagnosis, or treatment." You can argue that like, well, yeah, you can say that, but then if you're still giving medical advice, that is still problematic. But I think, there are a bunch of other elements here that, are sort of hidden by the, the lawsuit and making it sound as though people were directly misled into believing that this was, real therapist in some

Ari Cohn 13:53

Yeah, and I don't think there are any allegations that anybody was actually fooled into believing. Like, this is Oh, that's bad because people shouldn't pose as licensed professionals when they are not," and that's all we need. I understand where people's concern for the potential harm might be, but we're not there.

Mike Masnick 14:13

Yeah. it, this does feel like so many AG-inspired lawsuits are, that this is one for the headlines more than for any actual protecting of the people of the state.

Ari Cohn 14:25

Yeah. And you know what? At the end of the day, when you are asking for a court order for the company to change the outputs of its, chatbots, that is going to be a First Amendment issue. And when we are, making those decisions and deciding whether or not that's constitutional, one of the things we ask is, is there an actual problem? Like, is, is there evidence of, of an issue of actual harm? And, sometimes, for better or for worse in our constitutional system, you have to wait for something to go wrong or, at least have evidence that it is going wrong, before you can use the heavy hand of the government to, you know, fix a, a societal ill like this. And, you know, I realize that's not entirely satisfying for a lot of people, but, there are guardrails on speech regulations the government can enact for a good reason.

Mike Masnick 15:16

Yeah. and again, don't wanna sound flippant here either in, in terms of

Ari Cohn 15:21

Right. Cause it could be

Mike Masnick 15:22

Yes, there certainly could be areas where it is serious, but, the flip side too, I do think that there are people who find value in sort of fictional chatbots that are sort of literary devices. and that's, to me, that's what this one clearly is, that it's, reading those examples, like it is clearly designed to be playing a role and not be seen as a, as a serious person. And I could see scenarios where that, sort of interactive fiction can be fun for some people. It's, you know, like people get lost in, books for a reason, and what if you could talk with characters in those books or, experiment in different ways have those kinds of interactions with an AI chatbot. It's not something that appeals to me, me personally, but like...

Ari Cohn 16:06

I'm sure I do a lot of weird things for entertainment that other people would, would find bizarre. So, you know, I'm gonna let people have what they, what they want.

Mike Masnick 16:14

Yeah. So it's, it's, it's interesting. I mean,

Ari Cohn 16:17

But the governments are coming for that too. You know, last session in California, there was that bill that would have prohibited any, uh, chatbot from prioritizing anything other than facts and reality, uh, or honestly, like robots, outlawing robot Santa Claus. Like, you know, if people wanna have AI friends, like let them, but the idea that we are going to keep people from that interactive environment... I mean, the Supreme Court has, said on multiple occasions that the more interactive literature is, the better it is, and what we're seeing here is the government saying,"No, this is, this is too good. This is too interactive. This is too convincing." And when you take that to its logical conclusion, you get to a pretty dark

Mike Masnick 17:02

Yeah. I mean, I, again, and I've made this point a bunch of times in the past too, like, I do think that there are legitimate concerns where people get so sucked into these things they no longer realize that it is clearly fake. And, there should be efforts made by everyone, mental health providers, the companies, whoever else, to try and recognize where that is happening and try and figure out what are the targeted interventions that can actually help. But this idea that, effectively all literary chatbots couldn't ever pretend to be a, therapist or whatever feels off.

Ari Cohn 17:35

Yeah, and you know what? I definitely understand, like you, the concerns with people perhaps taking AI medical information with not enough grains of salt. Totally get why that's concerning to people. Some people, though, have found that it is useful for them to have a back and forth dialogue about medical issues they're experiencing, and it's not always harmful and it's not always helpful. You know, this is a very, depending on your circumstances type of thing. And, you know, do we really want to strip people of the comfort that they have found in having these resources which they would never have had before? I mean, This is definitely, I would say personally, I think normatively better than just having WebMD to freak you out, you know? At least I actually think that properly, uh, instructed, an AI can talk you off the ledge that that kind of WebMD, cycle of doom, can, you know, put you in. You can, you can get talked off of that, and I think that's a, a good opportunity for people who, might tend to freak out, So, you know, there's, there's benefits to it and we don't wanna, throw the baby out with the bathwater here.

Mike Masnick 18:42

Yeah. Yeah. Well, I actually, We didn't have this on our list, but I had a story on Techdirt this week actually talking about a story that last week's guest host, Jess Myers, had, had written looking at some of the research on that and finding that there is evidence that for many people these tools can be useful, in dealing with mental health issues. And it's, it's clearly not for everyone. And you can complain that, like, people shouldn't be turning to these, tools for mental health help, but the reality is that they are. And at that point, the question is then what do you do about that? And, if there are examples where it is helpful, learning about that and figuring out how do we, increase the helpfulness and, and in the cases where it's not helpful or actively harmful, figuring out what those are and figuring out how to identify it. But, like, the rush to just sort of say, y- No, ban it, stop it," I find to be, troublesome.

Ari Cohn 19:33

I don't put much stock in, you know, the"you shouldn't be doing this," argument. I think there are people out there who have much fewer options than, other people, and those people are trying to work with what they've got. And far be it from me to say,"You shouldn't be doing that." I think that's, uh, very myopic view of the situation here.

Mike Masnick 19:56

Yeah. All right. Moving on to our next story. there was an interesting study that came out. We talk a lot on the podcast obviously about age verification and the various child safety laws that are going into effect around the world, one of which being the Online Safety Act in the UK. And there was a study that came out this week from Internet Matters looking at the Online Safety Act and, uh, the title of the study is the provocative"Are Children Safer Online?" and it's actually pretty interesting and, you know, I think you can take from this report what you want in terms of if you are a supporter of the Online Safety Act, there are things in here that you can say,"Oh, look, it's working." And if you're not a supporter of the Online Safety Act, there are things in this report that you can look at and say,"Oh, it's not working." And so it is a, I think a mixed bag as I think as ever. right? to be honest about these things, all of these things are, are a mixed bag. A lot of this is survey-based too, which you can argue like surveys are not always the most accurate, way of collecting information. But I think the sort of headline is uh, there's a few different headline points that I, I, I took away from, from this study, which is that, kids and parents are seeing a lot more online age verification, online safety tools and, information. But they're also figuring out how to get past a bunch of the different checks. the headline of the article where I first found this report is the really striking one, which is that children are drawing mustaches on their faces to fool online age checks and it's working, which is, very evocative headline.

Ari Cohn 21:38

Incognito makes his return. There's a"Simpsons" reference for everything.

Mike Masnick 21:41

Yes. Yes. Uh, the study does quote one, parent suggesting that their kid had taken, uh, like an eyeliner pen and drawing facial hair. and the report does suggest that a few people had sort of said this was happening and that it worked, but it was more in the context of there appear to be a number of ways to get around online age verification tools, and people are using them. The thing that struck me as more interesting than just like draw a mustache on your face and you can get past the, the

Ari Cohn 22:12

Which is just terrific. I, I love it. No, no notes. Well done, children.

Mike Masnick 22:16

fact that, about a quarter of parents said that they've allowed their kids to get around the various age checks with, the majority of those actively helping them get around the age checks. and the, the rest being ones who just sort of turned a blind eye, that knew their kids were getting around the verification and were like,"Well, whatever."

Ari Cohn 22:36

Turns out parents want a little bit more control than the government saying,"You, this is the way it's gonna be."

Mike Masnick 22:42

And the thing that struck me about this is like this goes way, way back. Like I remember Dana Boyd over a decade ago, writing about COPPA the US law, which is why like a whole bunch of services technically officially ban kids who are under 13 because there are different rules. If you're targeted at kids under 13, you have different rules that you have to, do regarding data collection. And...

Ari Cohn 23:08

just to be clear, that's, uh, as opposed to, like, a mixed-use site like, ESPN where, kids under 13 might use it, but so do... You know, it's not aimed at them directly.

Mike Masnick 23:16

Right. and what Dana had pointed out through a bunch of, research and studies and talking to people was the real impact of this was teaching kids how to lie. That you know, there would be services that kids would wanna use and their parents thought was a reasonable use of the technology. Things like, getting on a chat app to be able to talk to grandparents is a classic example, and one that, I experienced with my family in terms of, like, getting my kids so that they can talk to my parents involved having to lie about their age. but it's, it's done in this way, as Dana pointed out, you know, 15 years ago, I don't even know how long ago it was, that, all we're doing is, like, teaching kids that it's okay to lie because the parents are coming in and telling the kids,"Oh, for this, it's okay to lie about your age." And now we seem to be doing that at an even greater scale where there's even more, tricky ways of, getting around this stuff. I don't find that to be particularly a good thing. I mean, I guess it depends on your view of, like, whether or not kids should learn how to tell lies. But I found that to be problematic. I, I will say that there is other stuff in this study that is interesting, including the fact that many kids and parents claim that they feel safer on the internet because of all of these technologies and, and tools. I wonder how much of that is, like, suggests that this is kind of security or safety theater, that, like, we have done something, therefore now the internet is suddenly safer. But, uh, what was your take on looking at this study?

Ari Cohn 24:53

You know, the, the part that I found actually the most interesting was the parts where both kids and adults notice changes in the tools available to help either parents control things or kids to control their experience more. And what that tells me is that those tools are evolving, and it seems like parents and kids both really like using those tools when they work well and when they're easy to use and when they understand them, that they would, in fact, prefer those things to somebody imposing a system on them. And I think A, duh, and B, that's, the lesser restrictive means that these cases at the Supreme Court always come down to that we could, you know, monkey with. I think companies have a long way to go still in terms of making the tools accessible, easy to use and understand. But, you know, Congress could pass a law tomorrow saying, Here's a bazillion dollars to educate parents, uh, about parental control tools." And surprise, they haven't. Well, they haven't been doing much of anything. But, you know- It's because that kind of bill doesn't really, get you headlines soundbites on the news. And it's not flashy, but it seems like that's actually something that parents and kids are both responding positively to. We can make, efforts to increase the availability and ease of use of these tools. I think that would solve a lot more problems than, the top-down ham-fisted attempts at, regulation end up causing.

Mike Masnick 26:29

Yeah. I think there's something to that. And, like, as a parent that has explored a bunch of, the parental control tools, I find them to be kind of a mess, and often unclear. and there's a, a layer of tech debt. I mean, I, I mentioned this whole idea of, like, teaching kids to lie to get on services, which then created problems whereas, like, you know, I, I'm not even gonna name any particular tool, but there was one where, we lied about my kids' ages, and then, some interesting parental tools came in place, but I, uh, we couldn't use them because the system thought my kids were 26 years old.

Ari Cohn 27:06

Wow.

Mike Masnick 27:06

it's like, well, what do you do now? and so, There are challenging things, but I do think, like, yeah. if there were more focus on, like, actually providing useful tools, and, and a lot of the companies keep evolving those tools and, like,

Ari Cohn 27:19

And they should keep doing so. They, they can do a lot more than they're doing, and we should, encourage them and pressure them to.

Mike Masnick 27:24

Yeah. Yeah. I think that is fair. The other thing I thought was interesting in this report, there were a couple things about, like, even though people are feeling safer and even though these, new tools are in place, there's a whole thing about how, oh, 49% of children still say that they had experienced harm online. And this is interesting in that there's a few different responses to this, at least in my mind, one being this idea that you're not going to eliminate harmful content and, like...

Ari Cohn 27:51

Also, what's harm?

Mike Masnick 27:52

Right. that was the second part Cause, like, that sounds bad, 49% of children said they experienced harm online, but then there is this list of what is the harmful content. And first of all, each one of them is a very low percentage. Like, the absolute highest percentage is 12%, of kids said they had seen violent content. But what is violent content? Like,

Ari Cohn 28:13

Yeah. Is that like a, a meme of somebody getting punched or

Mike Masnick 28:16

Ex- exactly. Right? you could see, like, oh, did you see, like, a car crash? Is that violent content? Is that harmful to the person watching it? I don't know. And that was, like, the highest level one, and, a bunch of the other ones are just, like, seen racist or homophobic or sexist content, and that's bad, but also is going to exist online in, yeah, in some form or another. And again, it's, like, a very small percentage of, people doing this. So it's like, it's one thing to just say,"Oh, they've seen harm," but when you start to dig into it, it sounds like

Ari Cohn 28:52

I could classify every time somebody's rude to me on social media, which is like every other post, as harm. Like, if I wanted to, like, depends on your, your sensitivity.

Mike Masnick 29:01

deserve that.

Ari Cohn 29:03

There's an argument for that.

Mike Masnick 29:06

Uh, yeah. So it is interesting. the one other thing in this r- report that I, thought was noteworthy, was it talks about VPN usage, and how with all the ways that people are getting around the various tools and the age verification and, and whatnot, how few of them actually mentioned usage of VPN. and the report presents this as sort of a victory because People claim, like,"Oh, these laws are going to be useless because everyone is just going to shift to a VPN, and it won't matter." And this report says,"No, it turns out that only a very small percentage of people actually used a VPN." which I, thought was interesting, but also probably kind of depend, again, like all this stuff, it depends on different things, And that includes, like, how restrictive the various tools, are. And so the...

Ari Cohn 29:58

whether people are accurately reporting what they're doing, you know. People might not want to admit they're using a VPN.

Mike Masnick 30:04

that the people who are using VPNs are not likely the people who will be surveyed by...

Ari Cohn 30:10

Yeah, most likely.

Mike Masnick 30:12

so there, there are a bunch of things there. But it, it is, it is an interesting report. It is, interesting that, you know, the big takeaway that they present is that people are feeling, both children and parents are feeling safer online, whether or not that's true, even as they're getting around restrictions, and even as they still claim that they're seeing harmful content online, they feel safer. which, makes me wonder how much of all of this is just kind of vibes-based anyways. the harms, the scariness, and then the, the laws to kind of fix it. But it is interesting to see some, some data at least on, on how all that's going.

Ari Cohn 30:43

Yeah, I think we're gonna have a lot more of that data in the coming years, and that's gonna play an increasing role in how we look at these things. I think that data would be very helpful, especially in court cases, but I think the court cases might move faster than the data in some cases.

Mike Masnick 30:59

I guess the, the one other interesting bit of survey data in this report was the fact that parents tended to see social media as much less safe than AI chatbots, which is on the list.

Ari Cohn 31:14

Which is so weird considering, like, when you read just news stories, like the general sense of the, the zeitgeist, uh, is that that's so bizarre to me.

Mike Masnick 31:24

Yeah. I kind of wonder if it's just, like, how long... Like, the nature of the tech lash and how long these different things have been out, and so, like, we've had 10 years of the media slamming social media as dangerous and harmful for kids, and AI chatbots have only been around for a couple years.

Ari Cohn 31:40

it's also, though, that, you know, this is something that I would point out when people were introducing bills, trying to prohibit AI chatbots from, quote-unquote,"grooming users." And like, an AI chatbot is actually literally incapable of grooming. It cannot effectuate the ultimate goal. Like, it just, it can't, it literally can't groom. And maybe parents are realizing, oh, if there's not another person on the other side, the chance of you running into, like, a bad person who means you harm is less for a chatbot, even if it poses other risks. Maybe that's part of the calculus. I don't know.

Mike Masnick 32:15

that's possible. I hadn't even thought of that, but yeah. I mean, social media is dealing with actual other people, and AI chatbots are not.

Ari Cohn 32:21

And other people are hell.

Mike Masnick 32:24

Yes. All right. Well, with that, I will move on to the next story, and this is... This one segues nicely,'cause we were just talking about VPN usage, and at least according to this survey, fewer people are using VPNs to get around, age verification laws. But there has been talk in all sorts of, all different kinds of places, about the VPN problem. So people who are pushing for age verification laws, there've been a couple of proposed laws in different places saying,"Well, if people can just use VPNs to get around these laws, we have to stop the VPNs." And so there have been a couple attempts. The UK has talked about it, a few other countries, a few states in the US. I think Wisconsin had a bill, maybe Indiana,

Ari Cohn 33:07

Yeah, Wisconsin ended up scrapping the VPN ban provision

Mike Masnick 33:11

So basically they wanted to, ban VPNs. But this week, Utah put in place an amendment to their social media bill that requires age verification that does not technically ban VPNs, but in effect absolutely does. Do you want, you wanna describe what, the details are here?

Ari Cohn 33:31

Yeah., Basically what it does is it places liability on the social media platforms for not accurately determining whether a user is in Utah and expects the social media platforms to just know when somebody's using a VPN. And I don't understand what exactly they're supposed to do, even if they do know if somebody's using a VPN, because that actually doesn't tell you that they are in Utah either. So

Mike Masnick 33:56

but, but but Ari, d-don't, don't they-- They're tech companies. Don't they have magic wands? Can't they just wave the magic wand and say,"I know you're using a VPN and you're really in Utah"? Isn't that how

Ari Cohn 34:06

Big nerd harder moment. It's, I mean, it, it's very unclear, like, the only thing that comes to my mind for how a platform would even know is if they see a lot of traffic from one IP address, which would indicate maybe that it's a VPN server. But there are tons of VPNs out there using tons of different servers. There are going to be people out there who are using less trafficked ones, and how is a platform even supposed to deal with that? You're basically saying,"Okay, well, if this other person over whom you have zero control does a thing to obfuscate where they are, you're going to be liable." Which doesn't make any sense, is a big due process problem because you aren't able to know the thing that you're prohibited from doing. and you're, it, it's placing liability on you for the actions of people out of your control. Like, does Utah want everyone using a VPN to be treated as if they're in Utah? Do I, is Facebook gonna treat me like I'm in Utah'cause I use a VPN? It, like, it just, it's unclear what they're after, what they want, or whether they really thought this through. But I, it's just really unclear.

Mike Masnick 35:10

Well, I mean, when it comes to internet regulations, the idea that the, the, the legislators thought this through or understood the technology is kind of a big ask. but yeah, I mean, to me this is, another example of sort of someone presented them pointing out the problems of their earlier law, which required age verification for social media usage, and people said,"Well, people are just gonna get around it with VPNs." And rather than recognize that that pointed to a problem with the underlying law, they said,"Well, we'll just do this and tell the tech companies, you guys are techies, nerd harder, and solve this problem, and that will solve it." and so the lack of actual technical understanding of what is and what is not possible leads to this kind of nonsensical result. the thing I'm wondering is, like, well, what happens now? Because, because it is this sort of impossible to deal with law, does everyone just ignore it? Are there lawsuits? Does, you know,

Ari Cohn 36:13

I assume somebody's gonna file a lawsuit over it, and it just... You can't comply with it. There's literally no way to comply. Uh, it's just, it really, I can't see any way that that kind of law can survive. It's just, it's just literally saying you're liable and there's nothing you can do about it. And, I, I don't know, you know, what, what to do with that. It's just, it's so outrageously, inane that it's hard to actually find the words.

Mike Masnick 36:40

Yeah. Yeah. I mean, the thing that gets me about these kinds of laws too is that I think it, it really decreases respect for the rule of law, right? I mean, if, legislators are passing laws that are literally impossible to comply with, that are just fundamentally incoherent to anyone who understands the technology, it makes just respect for, anything that a legislature does much less, and I don't see how that's helpful.

Ari Cohn 37:07

Yeah. It, it's not, I mean, you, you don't see it because it's not there. It, it's just, it is absolutely harmful and it really just it's not a good look. it really, it's just you're passing these laws with no idea what you're doing and what are we sending people to these legislatures to do? What are, what are we accomplishing here when legislators are just making things up that don't make sense? I mean, it's, it's embarrassing

Mike Masnick 37:33

Yeah. Yeah. I mean, I sort of wonder, like, where does all this go? Like, right now, obviously, all the states are, pushing different kinds of, regulations, and it's just sort of this widespread,"We have to do something about the technology. The technology's bad. the federal government is completely useless," and therefore the states feel that they need to step in and, do something. But we never see... I mean, almost all of the plans are, problematic. and this is just sort of a really clear indication of that, and that it's technically incoherent.

Ari Cohn 38:05

never gonna be able to like, stop people from doing people things. I mean, the law is not a solution for all of humanity's problems, and it can't be because the law cannot fix or really control even human beings. I mean, the sooner we realize that, the better off we'll be.

Mike Masnick 38:21

Yeah. Well, if you can figure out how to do that, I would appreciate it.

Ari Cohn 38:25

I'll, I'll nerd harder.

Mike Masnick 38:26

There we go. There we go. Uh, moving on, there's really, really interesting and worthwhile read, and obviously all the, sources we're using we'll have in the show notes. but there was this great piece in Tech Policy Press this week, by Apar Gupta talking about, India's digital censorship infrastructure. And I found this to be just a really, really fascinating read and really worthwhile because it goes into great detail in terms of what things have happened in India that gave them the justification to put in place rules to allow the government to demand certain content to be blocked. And this is interesting to me. A-a-a-about decade ago, I had written a report on, like, different intermediary liability and different censorship regimes within different countries, and one of the, the actually the most permissive and strongest protection-wise was India at the time. That was about a decade ago. And ever since then, they have backslid, drastically to almost o- the opposite extreme. And it's sort of interesting to see how they came from... you know, there were some early court cases or court cases in the sort of early 2010s timeframe that effectively found that India had the sort of IT regulations effectively created a Section 230-like regime that, that said platforms weren't liable. That has changed drastically over, over the last decade. And the thing that struck me about this article was as he walks through the reasons, these reasons all sound familiar. There are things about child safety and national security and, oh, we need to protect values, Indian values. you know, it's the, usual ways that you wrap suppressive laws of, Oh, terrorists," and,"Protect the children." that's all there. And then it wraps up in the, this, like, fear about AI deepfakes. That is, like, the sort of, crowning achievement is like,"Oh, AI deepfakes are going to do all this damage, and therefore we need these effectively censorship laws to allow the, takedown of, content." And then the article gets into, like, so what is actually happening? And it turns out, surprisingly for, anyone who was born yesterday, but perhaps not for, for anyone slightly older, these laws have been used You know, they've been used more and more, and almost every example that they have, though it's secret, the, the takedown demands are issued in secret, so they don't have a full picture. But of the ones that they do have a picture of, it is used repeatedly to take down content that is critical of the ruling party in India. So, satire, things that are just critical of the government, it is a censorship tool and is used for censorship and is used by the powerful party in control to suppress criticism. does that surprise you, Ari?

Ari Cohn 41:26

It absolutely does not. That's where all of this always goes, and they just did it really quickly. I mean,

Mike Masnick 41:32

Yeah.

Ari Cohn 41:32

people have been freaking out about deepfakes for what? Six years now, and there's still very little evidence that it's this gigantic problem that people have said it's going to be. It's so easy to scare people about deepfakes and all the things that they could do that it's low-hanging fruit. It, it's a target-rich environment. and it just... point to the things that have happened because not many of them have actually happened.

Mike Masnick 41:57

mean, the, and the crazy thing is, I- if I remember correctly, I think one of the earliest"Control Alt Speech," episodes we had, we had a discussion about how politicians in India specifically were using deepfakes in actually creative ways, where it wasn't being used to fool people, but was saying, like, they were creating deepfake versions that were clearly labeled as deepfake versions to allow constituents to ask questions. You know? Oh, you want, you wanna talk to this politician, you can call them on the phone or, through an online service and talk to them.

Ari Cohn 42:31

It's like what Eric Adams did in New York, which, you know, far be it from me to praise Eric Adams, but, you know, he, used AI to kind of, uh, have his voice conveying messages in other languages, and people were so upset because they thought,"Oh my God, he's giving people the misimpression that he speaks so many languages," which, whatever. But, like, it is so much more engaging and you feel so much more included in the polity when you hear something in your tongue from the actual person in their voice rather than through a translator, that, like, I think the benefits of that outweigh any negative that comes from people thinking that Er- Eric Adams speaks multiple languages. Like, who cares? Like, that is just such a positive for democracy that, like, it, it baffles me that people got worked up about

Mike Masnick 43:16

Yeah. but I do think that, this story about what is happening in India just replays a story that, that has been shown over and over again, where it is very easy to justify laws that are for the suppression of speech to base it on whatever scary thing that is happening, whether it's national security, protecting the children, some new technology is doing something bad and is harmful. And over and over again, it comes back to examples of how that then becomes used by the powerful to suppress the less powerful or the marginalized. And...

Ari Cohn 43:49

It also makes things a lot cheaper for people who want to challenge established interests and creates more options for people, and the people in power do not like that.

Mike Masnick 43:57

Yeah. So I thought this was a really, really interesting story. It's just, yet another case study of where speech suppression laws, get abused in, in, problematic ways. And then I think we'll move on to our final story of the week on another issue that is, I guess, depending on how you argue it, could be a moral panic or might not be. but, there's all this talk of banning mobile phones in schools, that we've been told that is part of the problem. That is why the kids these days are not all right, if, if we're to listen to some folks. and there's now a study that has come out, a fairly large-scale study looking at whether or not the cell phone bans in schools have worked. this was actually, was a pretty interesting study. You know, it's still fairly early, but this was pretty, pretty large scale, and they actually worked with a company that provides these bags that, you can put your cell phone in and sort of locks it for a certain period of time. and so a bunch of schools are apparently using those. So when you arrive in school, you drop the phone in a bag. And so they looked at the results in those schools versus other schools. And those other schools often still had some sort of mitigation. They weren't just letting kids use cell phones, wildly throughout the day. but the thing that they found that was most interesting and then sort of the headline grabber was that the test results were not any better. the schools that banned cell phones, if it was true that this was causing, bad study habits or whatever, kids weren't learning, the test results did not reflect that. There was no improvement in the schools that had the full, start to finish of school cell phone ban. There were also, the report found, at least briefly, though it looks like the effect went away, that over time, initially there were more disciplinary problems in the schools that banned cell phones. but it was unclear why. It could've been kids getting around the ban, it could've been kids

Ari Cohn 45:55

There are children involved. It could be a bazillion different things. The variables are

Mike Masnick 46:00

it could be kids who were used to looking at cell phones all the time suddenly have nothing else to do and don't know what to do with their time and just start doing vandalism. I don't know what, what, what it might've been. But they said that, that effect seemed to decrease over time, so that, may have just been a transitory effect.

Ari Cohn 46:17

Turns out one size fits all isn't really a great solution.

Mike Masnick 46:21

yeah. I mean, the, the study did say that, teachers still liked it. they claim the students were paying more attention, that they had less issues to deal with because, the cell phones were banned. but I think it's, you know, similar to what we said with the other study about the Online Safety Act earlier in the podcast. it's a mixed bag. These are complicated issues, and there isn't a clear, Oh, this is definitely the thing to do." And a full cell phone ban has some benefits, but also is not this, grand solution that fixes everything.

Ari Cohn 46:52

Yeah, I mean,

Mike Masnick 46:53

about it?

Ari Cohn 46:53

I'm not particularly, worked up about cell phone bans in general. Like, if schools wanna try that, like, I've always thought that's not a wholly unreasonable thing to wanna try. I think there are probably schools that it works better at than others. I think, you know, as with anything else, it really just depends on particular circumstances. I, you know, I can see reasons why you wouldn't wanna ban kids from bringing cell phones at all. Uh, I think that might get into, you know, some problematic areas. So maybe, you know, I think the pouches thing or, you know, making kids kind of surrender them is if you want to, go that path, that's probably the way to go. But you know what? I'm an old person. I went to school without cell phones and, you know, I do think that kids probably miss out on some of the interactivity with your classmates that one might have had. Um, I think there is a strong argument for that. So, you know, I don't think it's, it's an unreasonable position. So, I'm in favor of schools trying, seeing whatever works best for them. I, you know, I'm opposed to the government kind of mandating, like California, you know, mandating that all schools ban cell phones. I think that's just not going to be the way to go. But, you know, schools will figure out what's best for their students, and we should let them experiment.

Mike Masnick 48:07

Yeah, that's sort of my take on it too. I mean, I think that, schools can come up with more flexible rules. I mean, I've talked about this before, like I am in California, I have kids in, in school and, literally last night as we're recording this, the school board for the high school had a, a meeting about how, how the hell are we gonna implement the, the new California rules. And I, I thought that the high school actually did a fairly good job of how they handled devices in schools, which was that, it's not like at the beginning of the day you drop it in a pouch, you can't access it till the end of the day, but that as you walk into each class, they put a board on the wall in every classroom that has little pockets, and so everybody puts their phones in the pockets. You don't have it, but if there's an emergency or you need it or there's a scenario where it is actually important to get it, they can. And so we have had that where there was a case where the school went in lockdown because there was a fear of an active shooter, which is a wonderful thing that happens these days in the United States. Uh, and we got a text from my child telling us what is going on and that they're safe and all this. And to me, that is a valuable thing that they had access to the phone in order to let us know that they were okay and they were in a safe spot. and so I think having the flexibility is really important but I do think it is important to recognize that, you know, as the study said, it's not some, secret key to solving all the problems of kids and devices and whatever. It's not going to suddenly make them into better students, whatever, you

Ari Cohn 49:37

Yeah, again, it's a human being issue, and you have to deal with human beings instead of trying to paper over it by, presenting it as a technology problem. You, you have to, you have to solve the human level issues, and if you don't do that, you're not gonna really get anywhere.

Mike Masnick 49:49

I, I think, I think you have just keyed in on the theme of today's podcast. These are all about human issues where, people are jumping to the idea that it's a technological issue with a technological solution, and the reality is, is not. That's such a good way to sum it up. I feel like we should just stop there.

Ari Cohn 50:07

Yeah

Mike Masnick 50:08

But Ari, thank you so much for, for taking the time and discussing all this with us. lots of good stories this week. Lots of, interesting things to discuss. and as always, we thank all of the publications that we cited today. They'll all be in the show notes, so please go check those out. there's a bunch of really good stuff. And a reminder again to check out the Patreon, patreon.com/ctrlaltspeech, Control Alt Speech. Uh, you can sign up there. You have a few more weeks before we raise the prices on the, top level of support. and so everyone, thank you for listening. Ari, thank you for being such a wonderful guest host. We'll be off next week and then back the following week.

Announcer 50:53

Thanks for listening to Ctrl-Alt-Speech. Subscribe now to get our weekly episodes as soon as they're released. If your company or organization is interested in sponsoring the podcast, contact us by visiting ctrlaltspeech.com. That's CT RL alt speech.com.