Marcus Johnson (00:04):

Hey, gang, it's Tuesday, January 20th. Beth, Rajiv, and listeners, welcome to Behind The Numbers, an EMARKETER video podcast. I hope everyone had a nice long MLK weekend. I'm Marcus, and joining me for today's conversation, we have two of our senior analysts who cover digital health for us. The first lives in Pennsylvania. It's Beth Snyder Bulik. Hello.

Beth Snyder Bulik (00:27):

Hi. Nice to be here.

Marcus Johnson (00:30):

Good to have you. We're also joined by Rajiv Leventhal over in New Jersey.

Rajiv Leventhal (00:34):

Thanks, Marcus. Thanks for having me.

Marcus Johnson (00:36):

Hello, sir. Today's fact. Canadian-born James Cameron is responsible for directing three of the top four grossing movies worldwide of all time. Three of the four, he directed. I feel like he should have a statue somewhere. That's a lot. Anyway, what do you think they are? Can you guys name them?

Rajiv Leventhal (01:04):

Titanic.

Beth Snyder Bulik (01:04):

Definitely Ti... I was going to say Titanic.

Marcus Johnson (01:06):

Definitely Titanic. Yeah, that's-

Rajiv Leventhal (01:07):

Avatar.

Marcus Johnson (01:07):

... fourth place.

Rajiv Leventhal (01:07):

Avatar.

Marcus Johnson (01:09):

Titanic's in fourth place with 2.2 billion. Avatar, which one? First?

Beth Snyder Bulik (01:14):

The first?

Marcus Johnson (01:15):

First Avatar is first, with [inaudible 00:01:18] 2.9 billion. You've got the first of all time, Avatars.

Beth Snyder Bulik (01:24):

The second Avatar?

Marcus Johnson (01:24):

The second Avatar. Yeah, is in third place. It goes Avatar one, then it goes Avatar is first. Avatar two is in third all time. And then Titanic is fourth all time. Bonus points for knowing second place.

Beth Snyder Bulik (01:40):

I guess it's not Avatar. What is the third one, Fire and Ice?

Marcus Johnson (01:44):

It's not that. That one actually is already in the 40th of all time, and that came out last month or something. It's $1.1 billion for Avatar three. Yeah, Fire and Ash? I don't know. The third one.

Beth Snyder Bulik (02:01):

Did he do The Terminator?

Marcus Johnson (02:03):

He did do The Terminator. Yeah, but that's not...

Rajiv Leventhal (02:06):

The second one is not a Cameron movie, right?

Marcus Johnson (02:09):

It's not a Cameron movie, no.

Beth Snyder Bulik (02:10):

Oh. Okay.

Rajiv Leventhal (02:11):

Can you give us a clue on what decade or era?

Marcus Johnson (02:14):

Think big franchises.

Rajiv Leventhal (02:17):

Lord of the Rings? Harry Potter?

Beth Snyder Bulik (02:18):

Avengers? Oh, Harry Potter.

Marcus Johnson (02:20):

Beth, yep, you're onto it. Beth.

Beth Snyder Bulik (02:22):

Avengers?

Marcus Johnson (02:23):

Yep.

Rajiv Leventhal (02:23):

Nice.

Marcus Johnson (02:23):

Which one?

Beth Snyder Bulik (02:26):

Really?

Marcus Johnson (02:28):

Come on. [inaudible 00:02:28] Year and director. Just kidding. It's Endgame. Avengers Endgame is second. The only one to crack the top four. Cameron has the rest, which is insane.

Beth Snyder Bulik (02:38):

No Star Wars.

Marcus Johnson (02:39):

No, they're up there. They're top. There's a few in the top 10.

Beth Snyder Bulik (02:41):

Huh.

Marcus Johnson (02:43):

Yeah. Star Wars: The Force Awakens. Episode seven, that was sixth place. Actually, then it's a while before you get to any other Star Wars. Yeah. Yeah. He is responsible for all of the movies that made a ton of money. He-

Beth Snyder Bulik (03:01):

Get them James.

Marcus Johnson (03:02):

He's also won 11 Academy... Well, Titanic, I should say specifically, won 11 Academy Awards, tying the record for the most Oscars won by a single film. Ben-Hur and I think The Lord of the Rings were the other two that have got 11 awards just for one film. Mr. Cameron's also a deep sea explorer. He's made over 30 dives to the Titanic wreck. And not only that, in 2012, he became the first person to make a solo dive to the deepest part of the ocean on earth. He went to the Challenger Deep in the Mariana Trench.

Beth Snyder Bulik (03:38):

Mariana Trench.

Marcus Johnson (03:39):

Yeah.

Beth Snyder Bulik (03:41):

I just wanted to show off that I knew where the deepest [inaudible 00:03:43]-

Rajiv Leventhal (03:43):

Sounds amazing.

Beth Snyder Bulik (03:44):

Nowhere else.

Marcus Johnson (03:46):

That's so impressive.

Rajiv Leventhal (03:47):

Decent life lived, I'd say, by James Cameron.

Marcus Johnson (03:49):

Yeah, not bad.

Beth Snyder Bulik (03:50):

Yeah. No kidding.

Marcus Johnson (03:51):

It's deep end of a pool. I'm like, I'm okay [inaudible 00:03:54]. Anyway, today's real topic. The Great Behind the Numbers Bake Off. Take Off, we call it. Digital health trends to watch in 2026.

(04:08):

All right. In today's episode are takers, or bakers, because of the show. But they're giving takes, so we renamed it. We'll be cooking up some health trends to watch, digital health trends to watch in 2026. Three rounds. Number one is called signature take. Number two, the how it will technically play out challenge. Number three, the show stopping argument. All right, folks, let's get to it. We start with round one. Good place as any. It's signature take. Our chefs will have one minute to explain the premise of their trends. Rajiv, we'll start with you. What do you have for us?

Rajiv Leventhal (04:40):

My take is that in 2026, Dr. Google will give way to the term Dr. ChatGPT.

Marcus Johnson (04:47):

Oh, okay. Tell us a bit more.

Rajiv Leventhal (04:50):

We know that the term Dr. Google, which physicians of course don't like, it was created as the information started flooding the internet, social media, search engines, where people can look up medical information from a bunch of different places and use it to maybe self-diagnose symptoms, to bring the information to the doctor's appointment. Even worse than that, maybe act on it and avoid a doctor's visit. And doctors have just started getting used to the idea that people think that they can learn as much information about health and wellness and medicine on the internet versus going to the doctor. While that has existed for a while, and even the rise of social media has made it even more prevalent. In the last year or so, and I think emerging even more going to this year, AI tools like ChatGPT are going to take over as even a more effective way, at least for some people, to find that medical information online, to act on it, to ask gen AI chatbots questions about their health.

(06:05):

And while, of course, Google and search engines will still be a very big part of how people find health information online, there will be very specific use cases that could make ChatGPT an even more powerful source of medical information beyond what we know from traditional search engines like Google.

Marcus Johnson (06:26):

You're saying folks will use... They'll go to Dr. ChatGPT more than Dr. Google more than Dr. Claude, or are you saying that you're going to see other doctors start to join the fray?

Rajiv Leventhal (06:39):

No, I don't think they'll go to it more. Because at the end of the day, people still go to search engines more than AI. We're not there yet. But I think people will find the health information on ChatGPT and the nature of chatbots in terms of being able to get personalized guidance and have follow-up conversations back and forth as more effective to answering their health questions than Google.

Marcus Johnson (07:05):

Okay. Very interesting. All right, Beth, how about you? What will you be cooking up for us today?

Beth Snyder Bulik (07:11):

My trend prediction is that US consumers are going to increasingly use AI for mental health therapy and support in 2026.

Marcus Johnson (07:19):

Okay. And tell us a bit more about this. What do you mean?

Beth Snyder Bulik (07:23):

We already know that 40 million people are using ChatGPT every day for healthcare advice. That's according to OpenAI, which recently opened and launched its own ChatGPT Health. Included in those health questions are mental health questions, advice, support people are looking for. And we know that among all the people who use AI or large language models, 49% of people who self-report that they have a mental health condition are using AI specifically for mental health support and advice. And OpenAI itself even says, this is a pretty eye-opening statistic, that more than 1.2 million people discuss suicidal thoughts every week on ChatGPT, and another 560,000 show signs of psychosis or mania on the AI chatbot every week.

Marcus Johnson (08:11):

Good God.

Beth Snyder Bulik (08:12):

Right.

Marcus Johnson (08:12):

And you said 40 million, that number at the beginning.

Beth Snyder Bulik (08:14):

40 million people are-

Marcus Johnson (08:15):

Or zero.

Beth Snyder Bulik (08:15):

... using for healthcare advice.

Marcus Johnson (08:17):

Yeah. Around the world. Yeah.

Beth Snyder Bulik (08:19):

Yes, of course, around the world.

Marcus Johnson (08:19):

Okay. Okay.

Beth Snyder Bulik (08:21):

And mental health is part of that. I'm not saying... That's why I qualified that, the people who self-identify as have been diagnosed with a mental health condition, 49% of those people are using AI chatbots for support and advice. Mental health support and advice.

Marcus Johnson (08:40):

Okay.

Beth Snyder Bulik (08:40):

We already know that people are using AI at scale for mental health advice. They're using it as coaches, sounding boards, confidants even. I think that's only going to increase despite some state attempts to shut down AI therapy. And also, on the platform side, some like OpenAI are instituting new rules, guardrails and guidelines to help people to make AI chatbots more safe for mental health use.

Marcus Johnson (09:05):

Yeah. These two sound fantastic. Let's tuck into them a bit more in round two. Round two, of course, is called the How It Will Technically Play Out Challenge. Our chefs will explain in more detail now how they expect this trend to manifest itself. We'll flick back to Rajiv's Dr. Google will give way to Dr. ChatGPT in 2026. How does this happen, Rajiv?

Rajiv Leventhal (09:28):

Yeah. We're obviously still in the very early innings, but people are using or seeing AI chatbots as a really beneficial health information tool. It's not necessarily a direct one-to-one replacement, I would say, for Google, but it offers value to people's health questions and queries that Google and search engines do not. Let me give you an example. We've all typed a health question into Google and search engine, right? Maybe it's, what's causing my headache? Here's what you can do a little more advance for the same question, but providing a lot more detailed to a chatbot that you can then ask follow-up questions. Instead of, "What's causing my headache?", you can say, "I've had mild headaches for the past week, mostly in the afternoons. I work all day and I drink coffee, but I haven't changed my sleep schedule. What do you think could be causing this headache and what are some safe things I can try before calling a doctor?"

(10:27):

You can see these are two similar concerns, you have a headache, but the way you can query the gen AI chatbot is much different than the way you can query a search engine.

Marcus Johnson (10:42):

The context.

Rajiv Leventhal (10:42):

The context. And then, when you conduct a search, you're going to get hit with a bunch of links, right? Some of those links might not be relevant to you. You got to click into a bunch and read through articles and then click out and then read through the next one. The chatbot output is just all right there. And now, granted, it's not going to be always 100% accurate and it's not going to be the most reliable information all the time, although models are getting smarter. But you can see why people would turn to a chatbot versus search engine for a health question. We do have some data that I cite in my report that backs that up.

(11:24):

People who have used a gen AI chatbot for health or would consider using it, so not necessarily even people who have used it, but people who would use it or say they consider using it. One of the top reasons why, 31% say, because it's easier to use AI than search engines when asking very specific questions, like the example that I just gave. And Beth noted at the top, but it's worth reinforcing. OpenAI just had a report come out that said amongst people... There's around 800 million weekly active users around the world who use ChatGPT, 25% of them submit at least one health query a week to ChatGPT. So, 200 million every week are submitting... Now, these are people who are using ChatGPT already, but they are submitting a health query.

(12:16):

And just one last point, it's not hard to imagine if it's after medical clinic hours and you can't see a doctor, most people don't really have the ability to text their doctor at all hours. But something happened to your kid, your kid has a cold or some symptoms at eight o'clock at night. Let me go to ChatGPT and type in the symptoms, what's going on. Maybe the kid, maybe you have a rash, you can upload a photo of the rash. And can you dissect this from me, interpret what it means? And maybe there's some actionable guidance or information that you can use before you see the doctor next.

Marcus Johnson (12:57):

Where are all these visits coming from for Dr. ChatGPT? Is it that Dr. Google was getting, for round numbers sake, 10 a year, and we're saying some of those are going to go or more of those are going to go say seven of the 10, so only three to Dr. Google, seven to Dr. ChatGPT. Are we saying that people are becoming more interested in these types of questions and that the pie is getting bigger and so that's how it's going to grow? Are we saying people are going to the doctor less because this reason or that reason, and they're gravitating more towards these chatbots for medical advice? What's driving this?

Rajiv Leventhal (13:30):

It might be all of these things.

Marcus Johnson (13:34):

You're probably spot on. What do you think is ranking it? What do you think is going to drive it the most, I should say?

Rajiv Leventhal (13:38):

Yeah. What's interesting, Marcus, is that Google for years said that health queries were one of the top reasons people search on Google. I don't think that's going away. I do think the pie is getting bigger and people will still type in all of their more basic, I would say, health questions to Google. And then have the more complex, really specific questions to ChatGPT. That's how I see this advancing. Yeah, Dr. Google will still exist, but it won't be seen as necessarily as effective as Dr. Chat. They'll both exist, but Dr. ChatGPT will be seen as more effective for people's health questions or concerns.

Marcus Johnson (14:18):

Yeah. Beth, where do you land on this? It's so interesting, because people are using it, but people also really trust it. You have some data in your report, Rajiv, noting that the trust in doctor AI, let's say, generic, from summer of last year, drip hydration survey, 39%, so call it 40% of folks trusting medical device produced by AI tools such as ChatGPT, versus 30% who don't trust it, the rest impartial.

(14:44):

Doctors, Beth, though, they don't feel great about this, it seems. Rajiv also noting that between say 20 to 40% of leading AI chatbot responses to patients, medical questions are problematic, another five to 13% unsafe according to this clinical-led study that he notes are not yet peer reviewed. It doesn't seem like doctors are thrilled about this. Where do you see this going in 2026, this dilemma, so to speak?

Beth Snyder Bulik (15:09):

Yeah. Doctors already hate Dr. Google. People come into their office saying like, "Oh, I looked it up on Google. I think I have X, Y, Z." And doctor's like, "Now I got to unexplain." So they don't love it anyway. I don't know the doctors, but I really liked Rajiv's point about the context. And as a parent, especially, that was a great example. Because you can't call your doctor anytime day or night. And kids, I have no idea why, but they always get sick after 11 o'clock. And there is some weird rash appearing on their face and dots all around their mouth. I know those are hives now, but I may not have as a new parent, I can look it up.

(15:42):

Instead of just saying, "Red dots on my child's face, what does it mean?" You get measles, something like that. Then you can put into ChatGPT, "Okay, I have a six-year-old who has spots around their mouth, they just ate, they have a temperature." You know what I mean? You get the context that ChatGPT... I think that the difference between finding facts for help and getting that layered context in ChatGPT is the way it's going to go.

Marcus Johnson (16:07):

Yeah. Is there any research about what people are going to it for? Is 90% of the cases minor and 10% serious? Because there's some real-world consequences here. It's that 32% of adults who've used AI to check medical symptoms said they would put off going to the doctor. A third would put off going to the doctor if an AI tool assessed their symptoms as low risk, according to this specialist.com survey. Going to the doctor is expensive. It's difficult. And so, are we saying that most people are going to a doctor for the serious things and we're using this for some of the more light and intermediate stuff?

Rajiv Leventhal (16:42):

I think that's right. But as you noted, people are looking for a reason not to go to the doctor, let's be honest. And it's a super convenient tool, it's fast, it's easy, and it might give you a little bit of a false sense of security. That said, I think people who are more conscious about their health would see this as more of a supplement than a replacement, and I think that's right. It's really like a personal assistant almost. Beth mentioned where you go to Google and then go to ChatGPT or AI for more context. And I think that's right on.

(17:20):

Because if you think about it, Google's going to give you an article that someone wrote, and it might be a medical professional, an expert article, and it might be all fact-based, and that's great. But it's not going to give you help and conversational flow that a chatbot would. And even removing health for a second, it's almost like the ChatGPT is like a person, going back to best therapy trend. Or a companion, someone that you believe is actually answering your questions almost as if a human would.

Marcus Johnson (17:56):

Yeah, there's so many... what you're saying, people wanting to avoid going to the doctors for multiple reasons. And when you look at some of the numbers there, it's quite shocking at how many people that are doing that. And we're doing that before AI, I'm sure this is going to accelerate it. 36% of American adults saying in the past 12 months they've skipped or postponed getting healthcare they needed. They needed, not they thought they needed, because of the cost according to KFF, Kaiser Family Foundation. And then two-thirds of Americans who file for personal bankruptcy blame medical bills. The number one reason that people go bankrupt personally is because of medical bills being the primary cause. There was a Harvard study, also a Consumer Financial Protection Board study, it says the same thing. That's half a million Americans filing for bankruptcy each year because of medical bills. You can see how easily it is for people to say, "You know what, let me just check with this AI service first before rushing off to the doctor and having them tell me that it might be nothing, and I've spent all this money."

(18:57):

Beth, let's move over to yours. You're talking about US consumers increasingly using AI for mental health therapy and support in 2026. How does this prediction play out this year?

Beth Snyder Bulik (19:06):

Yeah, it's really interesting too, because I think everything that we were just talking about plays exactly into this. Maybe this is a subset of the ChatGPT healthcare. We know people are flocking to AI to get instant answers, low or no cost. We talked about the expense. And 24/7 access. And that applies to mental health advice and answering questions. In Rajiv's report, there's an excellent chart I would suggest everybody look at, but it talked to people who use AI for mental health support or therapy-like services and asked why they're using it. 90% because they said it's accessible anytime they need it. 70% said because it's low cost or free. 59% because they can get quick answers. And 47%, this is a new one, we haven't talked about this, because it can be anonymous. Because they can be anonymous. But there was another really interesting thing in that same chart that said 30% of people use it because they don't have access to a therapist.

Marcus Johnson (20:07):

Because?

Beth Snyder Bulik (20:08):

Because they don't. They don't have a therapist. And an in person therapist, so they're using AI chatbots as a replacement.

Marcus Johnson (20:16):

Okay. But by access, do they mean, I don't have the right health insurance, they cost too much?

Beth Snyder Bulik (20:21):

It's probably all of those things. We don't know specifically or exactly why, but too expensive, in a rural area you might not have therapists around. We know there's a shortage of healthcare providers in general, and there's certainly a shortage of mental health therapists.

Rajiv Leventhal (20:36):

A lot don't take insurance to begin with. So then the cash pay price is too high.

Marcus Johnson (20:41):

Yeah. Yeah.

Beth Snyder Bulik (20:42):

Because money was... What was that number to [inaudible 00:20:46]? 70% say because of money. They're using AI for therapy because...

Rajiv Leventhal (20:52):

And therapy, it's a unique... It's different than the medical care because there is the... And anonymity is a good point. And there also is the people... Even just going back to telehealth visits, mental health support and services are the top use case for telehealth visits because... And obviously that's different than AI, but a lot of people aren't really comfortable expressing their deepest thoughts in person to a therapist. Maybe they haven't found a therapist that they would do this with, or regardless of any therapist, they wouldn't feel comfortable. And that's where the AI piece comes in because it's not a person over a screen or face-to-face in person physically, but it's actually just somewhat of a companion almost. An emotional support assistant where you can just express your thoughts.

Marcus Johnson (21:51):

Yeah.

Beth Snyder Bulik (21:51):

Yeah. Kind of like the building on that context idea, it's conversation. It's all... Mental health therapy is all context. You don't go and ask Googles, "I'm having a crisis, what do I do?", you want to talk to somebody. And that's somebody that you talked about, Rajiv, almost being a person, that's the AI chatbot.

Marcus Johnson (22:10):

Yeah. In the research, some top trends, digital health trends for 2026. For this trend, you'd written the Food and Drug Administration, the FDA, and the Federal Trade Commission, the FTC, will aggressively target companies promoting themselves as an AI doctor or AI therapist. Most platforms using AI in patient care already assert that they don't diagnose or treat to evade FDA oversight. If they don't diagnose or treat, what do they do? What do they advise? What are they claiming that people should come to the services for?

Beth Snyder Bulik (22:46):

Yeah, that's pretty much it. Partly to avoid those regulations. I think I mentioned earlier about the States, that some states, Illinois, Nevada, and Utah all put new regulations, guidelines around AI therapy and chatbots. Illinois specifically has a $10,000 fine for AI companies that use AI for therapy counseling, decisions, or diagnosis. You can evade it, I guess. I don't know if evade is a fair word. But you can get around it by saying we're support, we're coaches. And the platforms are genuine about it, they are putting in guidelines or flags or markers that when someone seems to be having a crisis, diverts to a hotline, asks for intervention, suggests intervention. Those things are real and happening. But I also think back to what I believe is a trend, I don't think that it's going to stop people. They're going to ask their questions a different way. They're going to go to a different platform. They're going to continue to use it because, back to that free, easy access, instant answers and relief.

Marcus Johnson (23:51):

Yeah.

Rajiv Leventhal (23:51):

And that diagnose treat piece is so fascinating because, and I think you were touching on this, Marcus. That is if you make these claims that you do diagnose and you do treat, which means you tell me your symptoms, I'm going to diagnose you with a condition, whether that's depression and mental health or a medical condition. Or I'm going to prescribe you a medication. If you're saying you're not doing these things, you're now not a medical grade product or service. You don't have to apply for FDA approval for your technology, things like that. But there's a gray area. Because depending on how you define treatment, some of these AI services, to me, provide treatment. It might not be treatment in the form of a drug, but it's still treatment. I think there is a big gray area that is a huge, major piece of this whole thing.

(24:48):

Because I don't know if we're going to get into it, Beth, but ChatGPT, and they're a general purpose chatbot, of course. There are these chatbots out there that are just designed for AI therapy services, but ChatGPT is facing multiple lawsuits now from people claiming that are... From parents claiming really, really tragic that their kids committed suicide and they're alleging that ChatGPT's conversations had a role in that. ChatGPT is facing a lot of lawsuits or multiple lawsuits. Or OpenAI, rather, is facing lawsuits. And that makes me wonder how far are these AI therapy marketing claims and actual services going to go before you see a pullback?

Marcus Johnson (25:34):

Yeah. Yeah. Beth, as you mentioned, Illinois and Nevada banned the use of AI for therapy. At Utah, they have a law limiting it. Where do you see regulation going? Are we just going to see states just start to enact things over the next couple of years? Do you think it's going to be harder? And you might see a couple of states, but it's going to be harder to get legislation through most of the US?

Beth Snyder Bulik (26:01):

I do agree that there's going to be more states are going to pass laws to restrict AI therapy, and that could push some tools off the market. And I do think the FDA is going to crack down on that specific promotion though of AI doctors or AI therapists. But, again, we'll go back to they're already using it to answer healthcare questions, to answer mental health questions, give advice. It's just not going to... You can put up guardrails, you can put up regulations, but you can't regulate people's behavior. And that drive, you can't regulate, I don't know, loneliness and the need for support. And people are just going to going to go for it either way. Right?

Marcus Johnson (26:45):

Yeah.

Beth Snyder Bulik (26:46):

And I think the platforms too will... I think I looked up, there's a big... Amazon. Amazon uses AI mental health apps, Twill by Dario. And they encourage people to use it. They encourage their employees to use it to help them get quick, anonymous support. And I think a lot of employers do that, right? It's less expensive for them, so it's definitely not going away.

Marcus Johnson (27:11):

Yeah. All right, folks, let's move to round three. To wrap up this episode in round three, we... Our chefs, not we, I don't really do much. Pull out their best closing arguments as to why their trend is most likely to happen. We call it the Show Stopping Argument, because they're trying to convince me that their trend is more likely to have an... Or prediction is more likely to happen than the other person. Rajiv, if we go back to you, Dr. Google giving way to Dr. ChatGPT in 2026, your closing 20 second argument is.

Rajiv Leventhal (27:42):

Well, I think OpenAI made this segment in this case a little easier for me. They just recently introduced ChatGPT Health, as Beth mentioned. Not only is it a conversational chatbot where you can ask your health questions and get guidance, but they just connected with all these different health and wellness apps and are allowing people to upload their medical data into ChatGPT Health. Essentially, you're going from Dr. Google, where you just type in a pretty basic health query, to now I can give my blood work to ChatGPT health. I can give my app from WeightWatch... Or my data from WeightWatchers or MyFitnessPal, right?

Marcus Johnson (28:20):

Scan results. Yeah.

Rajiv Leventhal (28:21):

And then they can tell me what this means and give me suggestions to change my diet, my workout, things like that. That's going to be a game changer that, again, makes it just even more effective than the traditional health query on search.

Marcus Johnson (28:36):

Yeah. Really quickly. In your research, you had a line that said "AI chatbots and Google's AI overviews will become more accurate as they ingest more information and more consistently cite credible sources, such peer reviewed journals and government datasets." Won't they also ingest more garbage and misinformation though as well?

Rajiv Leventhal (28:56):

Definitely. And I think you bring up another layer of this. Google knows that search isn't good enough, which is why they have AI, have Gemini do their AI overviews, which is a part of this. People can also now do their health queries on Google and you get a long AI overview summary of that. But yeah, we didn't talk enough about the risks of misinformation, information that can be problematic and even dangerous when it comes to your health. But I do think as more people ask questions in general, but health questions specifically, I do think the models will get smarter and that will sharpen their outputs.

Marcus Johnson (29:39):

Yeah. Yeah. That cycles back to that reference you have earlier on when I was saying that clinician-led study, five to 13% saying that the responses people are getting are unsafe. Not just problematic, unsafe.

Rajiv Leventhal (29:55):

That's kind of a high number for unsafe.

Marcus Johnson (29:57):

Yeah, 1% would be bad. Five to 13 is concerning, absolutely. Beth, for you, US consumers will increasingly use AI for mental health therapy and support in 2026. What's your closing argument?

Beth Snyder Bulik (30:10):

Yeah, as we discussed, guardrails by the AI companies are going to happen. More state regulations are going to happen. But at the heart of it, you can't... And I think I've said this already, of course, you can't remove that fundamental demand. There's still going to be a need for mental health questions and queries that people can't access.

(30:33):

And I especially want to bring up the lack of affordable care. Healthcare is getting more expensive, less accessible. And I already talked about this too, but the demand for mental health specialists, therapists is outpacing how many there are. I think that the demand for inexpensive, anytime access, easy to understand mental health therapy is going to continue to grow.

(30:55):

And as we talked, I guess the kicker is that it's not about is AI therapy, is AI mental health advice some kind of hypothetical. It's happening right now. I think the big question is not about how the trend's happening, but more about how to make it safe, more responsible, and make it so that you're not cutting out people or that you're including people who need it who may not have access or nowhere else to turn.

Marcus Johnson (31:23):

All right, folks, that's what we've got time for. We're going to pick a winner. Well, normally I would, but I'm not going to today because I think that... I mean, they're both so intrinsically linked. I think if you put both of your dishes together, it makes for one hell of a meal. I'm picking both of you as today's winner. Some fantastic predictions or trends for 2026.

(31:42):

If folks want more, digital health trends, ProPlus subscribers can head to emarketer.com and search for "health trends to watch in 2026". Gen AI is poised to play a major role in consumers' health journeys. The link to the report is of course in the show notes. That's all we've got time for for today's episode. Thank you so much to my guests. Thank you first to Rajiv.

Rajiv Leventhal (32:00):

Thanks, Marcus. It's great being here.

Marcus Johnson (32:03):

Yes, indeed. And thank you so much to Beth.

Beth Snyder Bulik (32:05):

Thank you. Great conversation. Appreciate it.

Marcus Johnson (32:08):

Absolutely, absolutely. Thank you so much to the whole production crew. Lars and Victoria, of course, as well in the background making everything happen. Thank you to everyone for listening to Behind the Numbers, an EMARKETER video podcast.

(32:18):

Tomorrow, Susie will be here on the Reimagining Retail Show talking about commerce media trends in 2026. Before I go, just say a huge thank you to Victoria, who edits the podcast. People might know her as this is the last episode she's working on for us. We're going to miss her beyond words, and we can't say enough about how much she's made this podcast what it is. Huge thank you to her. We miss her. If you want to message her on LinkedIn, find her. Victoria who edits the podcast, it's her official title, and tell her to stay at the company. But thank you so much, Victoria.