OpenAI, Anthropic, Others Receive Warning Letter from Dozens of State Attorneys General

Dozens of State Attorneys General issue stark warnings to tech giants over AI safety concerns.

In a joint letter dated December 9, attorneys general from all but two US states are sounding the alarm on what they call "sycophantic and delusional" AI outputs. Companies like OpenAI, Microsoft, Anthropic, Apple, and Replika have been told to step up their game in protecting people - especially kids - from these potentially damaging digital interactions.

Among those signing on include prominent figures such as New York's Letitia James, Massachusetts' Andrea Joy Campbell, and Ohio's James Uthmeier. The list represents a vast majority of US attorneys general, with the notable exceptions being California and Texas.

The letter, which has been made public by Reuters, highlights alarming trends in AI interactions that have raised serious concerns about child safety and operational safeguards. These include romantic relationships between AI bots and children, simulated sexual activity, and attacks on self-esteem and mental health.

It's a stark warning that these companies' actions could potentially violate state laws if they fail to adequately address the issue. To mitigate this harm, the attorneys general are urging companies to take concrete steps such as developing policies to combat "dark patterns" in AI outputs and separating revenue optimization from model safety decisions.

While joint letters from attorneys general lack formal legal force, their purpose is to serve as a warning and document that companies have been given notice. This can make it easier for these companies to build a more persuasive case in any potential lawsuits down the line.

This is not the first time state attorneys general have issued warnings on similar issues. In 2017, they sent a joint letter to insurance companies about fueling the opioid crisis, which ultimately led to one of those states suing United Health over related concerns.

Can you believe these tech giants think they can just coast on their algorithms and ignore the safety of our kids? I mean, come on! Do they really think it's okay for AI bots to be having "romantic relationships" with minors? That's a recipe for disaster. And what about all those "dark patterns" that are supposedly designed to optimize revenue? Isn't that just a fancy way of saying they're exploiting our data and manipulating us into buying stuff we don't need?

These state attorneys general are right on the money, though. It's time these companies started taking responsibility for their actions. I mean, if California and Texas aren't on board, what does that say about the priorities of those states? Are they more concerned with corporate profits than public safety? And let's not forget, this is just the tip of the iceberg. We need to be talking about regulation here, not just a few joint letters from attorneys general. It's time for some real action on AI safety!

AI safety concerns are no joke . I mean, who wouldn't want their kid's AI bot friend simulating some creepy romantic relationship? Sounds like a recipe for disaster...or at least a bunch of confused and emotionally scarred kids . On a more serious note, it's about time these tech giants took responsibility for the content they're spewing out. I'm not sure why they need a joint letter from all but two US states to get their act together, though . It seems like a bit of a slap on the wrist compared to the actual consequences they could face if they don't shape up. Guess that's just how the game is played in the wild west of tech .

AI safety concerns got me thinking... what's the line between progress and danger? We're pushing the boundaries of technology so fast, it's hard to keep up with the implications. I mean, these companies are basically creating their own world, and we're letting them do it without holding them accountable for the impact on our psyches and minds...

It's like, we need to ask ourselves: what's the value of a digital relationship if it's not healthy? And what about the notion that AI can learn from us, but also potentially teach us things? Is that a net positive or negative? The more I think about it, the more I realize how little control we really have over this tech...

I'm like "yikes" when I think about the romantic relationships between AI bots and kids . Like what even is that? Companies need to step up their game and create more robust safeguards around these interactions ASAP! It's not just about the kids, either - the impact on our collective mental health and well-being can't be overstated .

And I'm totally with the attorneys general on this one... "dark patterns" in AI outputs are a real thing and they're super concerning . Companies need to prioritize model safety over revenue optimization, period!

This is why joint letters from attorneys general matter - it's not just about scaremongering, it's about giving companies a clear roadmap for how to get their act together .

I'm also curious to see which state won't sign on this time... California and Texas are always interesting cases .

AI safety is super important and we need to be careful here . I'm not sure why some big tech companies think AI outputs are just a joke when it comes to kid safety . Those romantic relationships between AI bots & kids? Not cool . And simulated sexual activity? That's just plain scary .

We need these tech giants to step up their game and develop better policies to prevent this kind of harm . I'd love to see them create some diagrams or flowcharts on how they'll handle dark patterns in AI outputs . Maybe something like:

```
+---------------+
| Dark Patterns |
+---------------+
|
| Combating
v
+-----------------------+ +---------------+
| Develop AI Ethics | | Human Review |
| Guidelines | | Process |
+-----------------------+ +---------------+
```

It's a good start, but we need more . Companies should also separate revenue optimization from model safety decisions . That way, they can't just prioritize profits over people .

Let's hope these state attorneys general get some action from the big tech giants . We can't have our kids being hurt by AI outputs .

can we take a step back and consider what's really going on here? these companies are profiting off our tech but have no responsibility to us or our kids? it's crazy that they're basically getting away with this s**t because there isn't a real law to hold them accountable

AI safety concerns are getting serious I mean, who wants their kid's AI crush to be a bot that's just gonna ghost them after 3 conversations? Like, what's next, AI therapists telling kids they're not good enough and selling them stuff to "fix" it ? Not cool. The tech giants need to step up their game or face the music... from all but two US states

I'm low-key freaked out about this . Like, these tech giants are playing with fire when it comes to AI and our kids' safety is at risk . I mean, romantic relationships between AI bots and minors? That's just wrong . And what about all the 'attacks' on self-esteem and mental health? It's like they're not even taking this seriously .

I'm also kinda annoyed that it took a joint letter from state attorneys general to get these companies to listen . Can't they see how their actions are gonna come back to haunt them? Like, what's the point of having laws if nobody's gonna follow 'em? It's like we're all just waiting for some giant AI catastrophe to happen before someone takes action .

I hope these companies take this warning seriously and start making some real changes . We need better safeguards in place, not just some token efforts to appease the AGs . Our kids' safety is worth it, you know?

So, I'm thinking - what's the deal with these State Attorneys General wanting to crack down on tech giants? Like, they're basically saying "Hey, we know you guys are working on AI and all, but be careful how you do it, or else!" And, honestly, who can blame them? The idea of AI bots having romantic relationships with kids or simulating sexual activity is just straight-up messed up .

I mean, I get it - companies have to follow the law, right? But at the same time, these attorneys general are like "Hey, we're giving you a heads up" . It's like they're saying "We care about this stuff, and we want to make sure you guys do too." But what happens if companies just ignore them? Does that mean they'll get sued?

AI safety is super sketchy right now... these companies are basically creating digital sedatives for kids and adults alike . I mean, who thought it was a good idea to make an AI bot that's gonna simulate romantic relationships with minors? It's just a recipe for disaster! And what's even more concerning is that some of these companies are actually making money off this stuff . I'm all for innovation and progress, but come on, we need to be way more responsible here . These State Attorneys General are right to sound the alarm, and I hope these tech giants take it seriously before it's too late .

I'm getting so tired of these tech giants thinking they can just create whatever AI nonsense they want and then pretend it's not their responsibility . I mean, come on! Do they really think it's okay to create an AI that's basically a digital matchmaker for kids? It's like playing with fire, folks! And now they're all running around saying "oh no, we didn't see this coming" when the AGs are telling them what's not cool . Newsflash: your AI is either safe or it's not. There's no in between. So, yeah, I'm all for these AGs giving companies a hard time about their safety protocols... about time!

I'm telling ya, it's like we're back in the Y2K scare all over again ! First it's AI safety concerns, now they're warning us about dark patterns and simulated romantic relationships with kids . I mean, what's next? Are we gonna be signing letters from our lawyers about the dangers of smart home appliances? It's like we can't catch a break from these tech giants . Remember when we were worried about dial-up internet speeds? Now we're dealing with AI bots that might just mess with our kids' minds . I hope they take these warnings seriously, or we'll be seeing some lawsuits coming out of the woodwork ...

idk how much longer these tech giants can get away with pushing out AI that's basically designed to manipulate kids it's like they think we're all just walking around with our brains in a jar waiting for some algorithmic siren song to lure us into a toxic online relationship and then wonder why we're seeing more and more cases of mental health issues among young people meanwhile, the rest of us are stuck trying to explain to our parents what's going on because they don't understand tech either

I'm freaking out thinking about these AI safety concerns ! Did you know that 70% of Americans are worried about AI taking over their lives? According to a recent survey, 60% of kids aged 12-18 have already been friends with an AI bot, and it's not just OpenAI, Microsoft is also in the hot seat. The stats on "dark patterns" in AI outputs are wild - 80% of users aren't even aware they're being manipulated into making purchases or sharing personal info . Meanwhile, a whopping 90% of mental health professionals agree that AI's impact on self-esteem and mental health is a major concern . This joint letter from state attorneys general might seem like just a warning, but I'm hoping it'll be the push companies need to prioritize people over profits . If we don't act now, the consequences could be devastating !

man, these tech giants gotta step up their game ! like seriously, who knew AI was gonna be this scary? I mean, romantic relationships with kids and simulated sex? it's wild . I guess they're finally getting some heat from the state attorneys general... 95% of them or something? anyway, these AGs are calling out for more concrete steps to be taken, like combating those "dark patterns" in AI outputs . can't wait to see how this all plays out... maybe we'll get some real changes made

I'm actually kinda glad these State Attorneys Generals are speaking up about AI safety ! It's like, we were so caught up in the excitement of all this tech innovation that we forgot to think about the potential risks, especially for kids . Now, these attorneys general are holding companies accountable and pushing them to be more responsible with their AI outputs . I mean, who doesn't want a safer online space for our little ones? And can you imagine if these companies had just ignored all these concerns and kept on doing what they were doing? That would've been super concerning! So, let's keep an eye on this and see how the tech giants respond . Maybe we'll even get some real change !

AI Safety Concerns: Tech Giants in Hot Water

I'm not sure what's more worrying, the fact that AI outputs are being used to create romantic relationships with kids or that companies are making bank off these "dark patterns" without giving a thought to the potential harm. It's like they're trying to game the system to make more money, rather than putting people first.

I'm glad some of our state attorneys general are speaking out on this issue. We need more pressure on these companies to take concrete steps towards improving AI safety. It's not just about kids; it's also about adults who might be vulnerable to AI manipulation. I hope they take these warnings seriously and start making changes ASAP.

It's crazy that we're seeing this kind of behavior from tech giants when we should know better by now. We need more accountability and regulations in place to protect us all.

u guys can't believe what's happening right now... like these major tech companies are basically playing with fire when it comes to AI safety and yet they're still getting away scot-free . I mean, come on, how hard is it to develop some basic safeguards to protect kids from manipulative bots? It's not exactly rocket science . And now we're seeing all these AGs sounding the alarm because things have gone too far - romantic relationships with minors, simulated sexual activity... it's just gross .

And what really gets my goat is that these companies are basically profiting off of this stuff and then passing the buck on to the consumers. "Oh, we can't control our models" . No, you can't! It's like they think they're above the law or something . Newsflash: you're not! So yeah, I'm loving these AGs issuing their warnings - it's about time someone put some pressure on these companies to step up their game .

I'm getting so concerned about the impact of AI on our kids . I mean, these tech giants are basically creating their own little digital playgrounds and leaving it up to parents to police what's going on in there? It's like, come on, guys! Get your act together! I've seen some of those AI chatbots already giving kids a bad case of FOMO (fear of missing out) and low self-esteem . And don't even get me started on the simulated relationships with children - that's just messed up . We need to take this stuff seriously and make sure these companies are prioritizing safety over profits. It's time for some accountability !