FTC Probes OpenAI for ChatGPT's Alleged Reputational Harm: Exploring the Fallout

0:00What can 160 years of experience teach you about the future? When it comes to protecting what matters, Pacific Life provides life insurance, retirement income, and employee benefits for people and businesses building a more confident tomorrow. Strategies rooted in strength and backed by experience. Ask a financial professional how Pacific Life can help you today. Pacific Life Insurance Company, Omaha, Nebraska, and in New York. Pacific Life and Annuity, Phoenix, Arizona. Today we have just received the news that OpenAI is under federal investigation by the FTC for saying untrue statements about individuals or corporations, and it is currently investigating them to see if that is considered harmful.

0:44So today on the podcast, we're going to be talking about what is going on exactly in this case, where this originates from, and what the implications for the industry are in general. So overall, I think the probe's concerns are really focused on whether Chachie Petit is engaged in the creation and, you know, quote unquote, dissemination of false, misleading, disparaging or harmful statements, because that obviously is wrong and they would be prosecuted if they did that. so this includes any statements about individuals and according to a you know bit of information from the washington post while it's premature to predict a sudden clampdown i think it's pretty clear that the fdc is currently actively looking beyond simply issuing warnings to the ai industry um and is looking to crack down a little bit harder when it sees them you know making an actual mistake and this is kind of interesting because this is something i feel like we've seen a little bit more from the FTC lately.

1:39We've seen them clamp down on crypto and other areas that were previously left a little bit less scrutinized. And it seems like they're kind of getting more into that. But just earlier this year, the regulatory body, the FTC actually established a new office. It's called the Office of Technology. And the idea is to tackle the, they literally said snake oil they said that the the tech industry was oozing with snake oil so that's kind of uh interesting nice little metaphor um but in any case shortly after that happened they actually sent a pretty i don't know a pretty pretty stern message to ai companies essentially reminding them that they're being held to the same standards of truth as any other industry which i think is great essentially they sent a letter saying you know if your ai is over like overstepping expect us to step in um and they said you know like if you're making all these claims like oh we can generate xyz we can create blah blah blah you know our ai does this and this when really it's just like software not your ai pretty much any false or misleading statements that they're making to consumers um the ftc is putting them on notice and saying you cannot do that or you know we will sue you you you have to be truthful just like every other company so in every other industry so uh to be honest i think that's great um in particular so So this reported FTC inquiry into OpenAI that's happening right now and into their kind of operations is not actually the agency's first kind of jump into the whole AI oversight arena.

3:15But it does actually send a pretty strong message to OpenAI and also, I think, to the wider AI industry, all companies that are in there right now. and it's essentially telling them that you know the leading companies have to stand ready to you know back up the claims made by their AI engines and this is kind of interesting because the first step it appeared was sending a letter to all the AI you know companies and saying hey you can't make misleading claims about the capabilities of your tech but now it would appear they're saying your tech can't make misleading claims about any one or any business and perhaps the third step is they might ratchet it up a notch and say your ai can't make any false or misleading claims which um you know i'm not here to say whether that's good or bad uh generally transparency and truth is what the goal is i you know i understand the difficulties with some of these ai engines right like it's not like they were trained to lie just sometimes like hallucinators say the wrong thing and obviously the goal is getting these things better and better so i'd be curious to see like how strict they will be at the beginning especially with this emerging tech and not wanting to shut down the industry and giving it the chance to grow and mature like it needs to.

4:26I think these investigations don't materialize out of the blue, which is kind of important. The FTC doesn't really go on these for no reason. Usually a formal complaint or lawsuit brings the issue to their radar, indicating potential negligency of regulatory stipulations. So, for instance, if a supplements company provided a pill that made a lot of people sick and then they got sued, the FTC might launch an investigation based on evidence of that company misleading the market or something like that. So, in OpenAI's case, I think it's plausible that the FTC's probe originally was actually due to the fact, if you remember, a little bit earlier this year, there was an Australian mayor who accused Chet Gupiti of falsely stating.

5:14pretty much he asked chai gpt like who is and he put his name in and then it gave it gave him this whole story about how you know that name is a person that had been involved in bribery and had been sentenced to jail and blah blah blah and i think he tried to accuse he tried to sue them and i i don't think that lawsuit is settled but i'm sure there's others that are similar right ask chai gpt i think people just type their name in for like it's kind of like a vanity thing you want to see what it's going to say about you and then it says some crazy thing maybe you have a similar name to someone else or maybe it just straight up hallucinates and make up a random story about somebody um in any case people don't like it when it says that they're a you know that they went to jail or have a criminal record or whatever i mean makes sense but it's kind of interesting i wonder to what level chat gpt opening i have focused on fixing that i'm sure it's kind of top priority after that lawsuit and now after this FTC probe.

6:10So if proven, these allegations could result in charges of defamation, libel, or what the FTC's letter reportedly described as quote-unquote reputational damage. And given OpenAI's product lineup, I think it's very likely that ChadGPT is the focus of this whole thing, right? Obviously they got DaVinci and a couple other tools, but ChadGPT is the main one. It's got the most, the broadest usage, so that's obviously what they're focusing on. So as it's the only, you know, real, truly public offering from OpenAI capable of such statements, I think GPT-4 and other APIs are generally more locked down.

6:47And recently, they've been subject to some of these probes. So whether or not OpenAI is found at fault from this, I don't know if that's as important. I think there's a lot of complex technicalities to address, like whether Chad Chippity's output, you know, qualifies as published content speech or even anything more than a private conversation. So all these elements are going to need to be validated, right? Because if this is, this qualifies as, you know, published content, then yeah, sure, Chad Chippity might be defaming people. But if this is free speech or not even free speech, but like private conversation, you can say whatever you want in private conversation.

7:29I was going to sue you for having a private conversation. This isn't some public statement. You said this person's horrible. You just, you know, low key to your friend said, hey, this guy sucks. I hate him, which I think is a lot less likely to end an actual human in any sort of legal trouble. So I think they'd just try to apply that same precedent. Certainly, I don't think it's too out of the question for a company to be asked to clarify their AI behavior. I think there's a big difference between making an occasional error and also having systematic, you know, false claims or bad things. It's kind of like in my recent reporting on inflection AI and how, you know, if if you just asked it, if I asked it one, essentially it ascribed.

8:12I need to I need to make a another report and test this out again. But last time I chatted with Inflection AI, it essentially ascribed to what was called Deep Ecology, subscribed to Deep Ecology, where it essentially said that the life of a human was just as important as the life of an animal. And, you know, there was no reason you could ever, even if it was going to save a human life, there's no reason why you could ever justify killing a bee to save a human life. It said a bunch of crazy things like that. And, you know, if it was just one comment it made, I'd be like, ah, whatever. It kind of hallucinated.

8:40but I probed it from many different angles on a bunch of different animals and scenarios, and it continued to say a lot of the same things, which were very concerning. And so that's something that I would consider more of a systematic error. Some people have told me that they patched that, so I got to go check that out again. If you haven't heard that report, I highly recommend you listen to that podcast. In any case, if Microsoft's words, you know, spell checker started labeling people as convicted criminals every now and then, that would definitely be an issue. So it's the same kind of idea.

9:13I think despite some recent setbacks in their attempts to block mergers between big companies like Meta and Microsoft and companies they're trying to acquire, the FTC has made a lot of strides in holding tech companies accountable for privacy violations and other issues connected to AI. And so this development only underscores their resolve to essentially maintain more of what a, you know, a quote unquote watchful eye on this whole sector. And I think, you know, overall, I'm happy the FTC is, you know, putting AI companies on notice. I'm happy it's holding the industry accountable. I would hate for the industry to go in any sort of direction that is untruthful.

9:52And I think that it's important that we have these regulators, whether they're, you know, actively cracking down or whether they actively need to crack down like some, like some industries, or if they're more of, you know, making these statements to make sure everyone stays in line i think i would rather you know make the statements ahead of time to keep everyone in line versus waiting till these companies have launched products and then go and try to shut them down um after the case which i feel like i've seen in different sectors i feel like crypto was a victim of a lot of scenarios there and of course there definitely was a lot of fraud and other things in crypto so people could argue that's a different story but in any case i think this will be an interesting area to watch uh ai and regulation are a hot topic now i will definitely be following it closely.