Welcome, everyone. Thank you so much for joining us for this conversation about Moderna. My name is Courtney Breen. I am the U.S. pharma analyst here at Bernstein, and it is my privilege to have Stéphane Bancel here with me, the CEO and chairman of Moderna. He's been in this role for a decent amount of time as well and has seen Moderna through the ages and through the different eras of the company. I'm really excited to have an opportunity to dive into Moderna today, where Moderna's come from, and where Moderna might be going in the future. I also know that AI is a super important topic for all investors these days. We'll be hoping to touch on the impact and potential of AI in drug discovery and in operating some of these businesses.
I do also want to remind you that if there are other topics that I'm not planning on covering that you'd love to have covered in this conversation, please do send them through the Pigeonhole app. You'll find a QR code to be able to send them through. I'll receive them up here and can integrate them into the conversation. We want to make this as relevant and as impactful for everyone that's here. Without further ado, Stéphane, again, thank you so much for joining us here today. As I mentioned, you've been leading Moderna for a while.
I think since 2011.
Yeah.
It was back before there were any products in market.
Yes
back before we knew about COVID, back before then we had the COVID vaccines. A lot has changed between then and now.
What has fundamentally changed in how you're thinking about the opportunity for Moderna and how you're running the business of Moderna today?
That's a great question. Good morning, everybody. Thank you, Courtney and the team, for having us here today. Yeah, I've been CEO since the beginning. I was employee number two, technically. There was a scientist at Flagship, RVC, and the founders I'd recruited, and I joined and then started to hire the team. Of course, the job has changed a lot. If I look very pre-pandemic, we were really a research and early development company. We had quite a number of products in the clinic. I think my belief in mRNA has increased tremendously-
for a few reasons. First, nobody had got an mRNA product approved in 2018. If we take the IPO timeframe.
pre-pandemic, whereas now we have four products that have been approved by FDA, another one being under review right now. A lot of readouts for phase III in cancer and in rare diseases this year, so it's quite a different world. The team keeps on pushing the boundaries of the science of the platform. I think we're still in the very early days of what mRNA can do for patients. I've always taken a very long view on things, which is when we started the company, I used the biotech technology as what I was trying to achieve with mRNA, which is if you think about Amgen and Genentech when they started in the 1970s.
which is more than 50 years ago now, they had no idea they could do antibodies. They were working on human protein, it was not working well in the early days, then they made the first growth hormone, insulin, then they made more, then they made first antibodies. If you look at the biggest sales of products, as we know in pharma today, it's mostly antibodies than proteins, right? We've always taken that arc to say, Okay, if we invest in science, if we can scale the company, if we can build a true platform, we should be able to get a lot of medicine over time into many therapeutic areas. I used to spend a lot of my time before COVID on just research and early development because that's all we were.
If you look at where we are today, we are very different companies present around the world. I would say I spend maybe 25% of my time on the commercial business, which is sales, marketing, manufacturing, cost of goods, scalability, pre-launch of products like intismeran. Right now we're in the thick of intismeran pre-launch activity, marketing, medical, pricing, sales. Same thing for PA. We've recorded that on rare disease because the phase III is reading this year. I spend maybe around a quarter of my time on research and early development because, again, as I just said, I believe we're still in the early days of mRNA. I believe we are the only company at scale today because the landscape has changed a lot. CureVac used to be an independent company. It was the first mRNA company. They are gone. They've been acquired by BioNTech.
BioNTech is splitting in two. They just announced they're also closing down their manufacturing facilities in Germany. Sanofi bought Translate Bio. Translate Bio, they just announced they are not working on the flu program anymore at Sanofi on mRNA. The world has changed a lot, and I think we are the only company today at scale to be able to keep inventing new technologies. I need to spend a lot of time helping the team find the next INT.
I spend around maybe 25% of my time on AI today because I think it's an amazing opportunity to reinvent the business, whether it's science or it's just how we do work. The rest is around talent and a bit of time with investors.
Fantastic. Super helpful. You touched on so many things that we're going to dive into in this conversation. Perhaps one thing that didn't necessarily come up is as where you're spending time is some of the policy pressures.
Yes.
Last year was a year where I think every single biotech and pharma company perhaps spent an inordinate amount of time kind of dealing with the potential risks.
policy risks that might have landed at the doorsteps of the sector. We're still seeing changes play out at the FDA particularly. We no longer have an FDA director or commissioner.
We've got an acting-
individual in that role.
Yes
who's come from the food side.
How are you thinking about the changes occurring in the regulator, kind of impact of vaccine skepticism, but also is there going to be consistency and therefore how does that influence how you're making decisions in R&D and the choices you have to make upfront that perhaps then don't land in front of the regulator for a couple of years?
That's a big question. I would say obviously for the industry it is really important to have predictability. Every CEO said the same thing over the last couple of years. Because of a very long lifecycle of our industry into our investment and returns, we need predictability. We need to know when we invest in something, is that what's going to be reviewed, is it going to be reviewed in the right way, and so on. That's of course, really important. The good news, I would say, is that most of the teams do the work. I think it's important to differentiate at an agency like the FDA, what the Commissioner and the Head of the Divisions do, which is around policy about the future. Of course, that is really important for the agency to prepare for a future.
Not having roles filled is, of course.
a problem for the country, for Americans, and their health, if you think about five, 10 years from now. If you look at the product reviews we're having, whether it's a vaccine, whether it's a rare disease or intismeran, it's the working team that are in constant discussion with all working team, and those teams are pretty stable. There's been a bit of changes early last year, but thankfully for our programs, we have not been really impacted.
Yeah.
The key players in all our key programs have stayed through, since early 2025. I'm, of course, like anybody else anxious to see stability in the agency, to be able to have people in a leadership position to talk about, for example, AI.
How do we use AI to file for approval? How do we use AI to do science? How do we use AI to enroll clinical trial faster, to select site faster? There's a lot of things.
Yeah
that we can decide ourselves as how we run the business. Because we are a regulated industry, there are a lot of things we need the regulator approval. We need policy clarity on those topics. I'm hopeful because I'm optimistic. I'm an entrepreneur, I have to be optimistic. I would be in different line of business. I think those things are going to get resolved.
Again, they don't impact us short term, they impact us on more long term.
Yeah, absolutely. Super helpful. I do want to get back to AI, maybe I'll dive into some important data.
Sure
Particularly as we've got ASCO on the horizon.
Yeah. This weekend
The big oncology conference and more data anticipated later this year for your intismeran program. Just for those in the audience that don't know, this is what previously might've been called a personalized vaccine, is now being termed as an individualized neoantigen therapy, that can be used to reduce the likelihood of progression or disease in cancer states. You've obviously got programs ongoing for melanoma. We've seen some very long-term phase II data demonstrating durability-
of the outcomes and outperformance relative to standards of care like KEYTRUDA, which previously were big step-ups. How do you think about what we might see later this year? What is the goal in terms of what could be demonstrated for this product? How different might it be to what we've seen in the phase II for the product so far? Perhaps just a broader question, what could be the potential future for something like intismeran if you were thinking a few years further out?
Oh, wow. That's a lot of questions. I want to try to take them in order, and if I forget anything, I'm sure you'll keep me honest. We shared the top line of a five-year data in January of this year with a hazard ratio of around 0.5, which we are very pleased about because it was the same as what we saw at three years.
As you said, the durability of a response was extremely exciting to us, what it can mean for patient and what it can mean for the company. What also was really exciting to us as we looked at the data in detail is the P value.
The P value improved as well, which when you see a study where you have duration in oncology and improvement in P value, this is really a good sign. It doesn't mean it's going to be a great outcome, but it's really, really important key enablers to a great outcome in phase III. We're excited about this weekend. We have two things at ASCO. There's, of course, the full data set of a five-year data Monday morning, scheduled as a plenary session, oral session, at ASCO. We have an investor day at the end of the day to walk people through the data who don't have direct access to ASCO. This weekend, again, it's public information, there's a poster about the translational medicine of intismeran, which is really at the T-cell level.
A very detailed work is being presented this weekend ahead of the clinical data to explain how intismeran works.
We've already shared a little bit at ASCO before, including, I remember in 2018, showing in lung cancer how we could take the blood from people, from patients, sorry. Before we inject them the product, we had sequenced their DNA of their tumor cells, so we knew exactly what antigen we coded in our product. We took the blood before we had their first injection of intismeran. We look at their T-cell for those neoantigen, and then we did the same thing after four dose, and we show that T-cells were starting to recognize the neoantigen that we knew we coded in our product, and that we knew the patient T-cell did not recognize before starting intismeran. It was a great proof of mechanism.
The question we all had, which is when we saw the clinical data two years, three years, and now five years, we are very pleased to see, do you have any clinical impact? You could have T-cell activation, but have no clinical impact, and that's not a drug. We saw the mechanism. You're going to get more detail on Saturday, then we saw the data, now the duration of the data. Like anybody else, I'm blinded to the data. When we have data, you know of data pretty quickly, as we always do at Moderna. We never keep data wrapped in the data important for obvious reasons, we're looking for the data. I think the piece that is exciting is that we just announced recently with Merck a new phase III study in lung disease. It's our second one, phase III in lung.
Just to tell you, Merck and all believes in the science and the mechanism. It's in stage 1 lung cancer patient. If you're diagnosed with lung cancer in stage 1, you do not get a checkpoint today because of a side effect of checkpoints. As we know, checkpoints, when they work, are wonderful medicine, but they come with a heavy side effect profile. The standard of care today is surgery, sometimes chemo, sometimes no chemo, and monitoring the situation. Here the idea is to try intismeran as a monotherapy. Within the study we're also going to try intismeran plus KEYTRUDA. We are very excited. Personally, I'm very excited about trying intismeran monotherapy, because we know from the phase I study of intismeran, we showed that at ASCO several years ago, that we had response in several tumor type, including lung.
Because of what we understand the mechanism partly was going to be presented on Saturday at ASCO on melanoma. We really believe that if you go earlier in disease, you have, first, less burden of disease, and two, you have a stronger immune system. What is very clear to us is that this drug works by teaching your immune system what to look for your T-cell to go eat your cancer. We believe that having a very healthy immune system is very important, which is why earlier in disease, we believe is going to increase the odds of our medicines working well.
Fantastic. That's super exciting. I think there's one question of does it work in melanoma? The next question immediately.
Sure
is how well is it going to work in other cancers?
Yeah.
Do they have to be highly immune sensitive cancers, or can they be kind of a broader set?
Sure
kind of perhaps heading towards your gastrics and other cancers that aren't as immune sensitive?
Yeah. Like we in science, we have to run the experiment because you cannot know a priori. We are deploying a lot of capital, and so is Merck. Again, for those of you that are not close to the partnership, it's a 50/50 cost share now and profit share. Every time you see a study going, you know that Merck is paying half of it.
It's a capital allocation they're making compare a lot of other assets they have in their pipeline, of course. Look, as I just shared, we have shown earlier response of intismeran as a monotherapy without even KEYTRUDA.
in melanoma, lung, head and neck. With what we know of melanoma and the mechanism of action, we believe, I personally believe, strongly believe, if not, I would not be deploying the capital either, that where KEYTRUDA works, intismeran should help improve the response from KEYTRUDA. How much is it? Is it a 0.5 hazard ratio like melanoma? I don't know.
We have to run the clinical experiment to know. Could it be better in some, worse in others? Most probably.
The piece that I'm excited also about, as I said minutes ago, with our lung stage 1 disease, is to go earlier in disease. For example, I learned talking to key opinion leaders, world key opinion leaders in melanoma, that some patient stage 2 melanoma do not want a checkpoint because of side effects.
Could one day we go earlier in disease, assuming the phase III is successful.
of melanoma later this year? Do we want to go earlier in disease? Because that could be interesting, and could intismeran monotherapy actually become standard of care because you don't have a safety profile. I think what most people are missing about intismeran, and you have to do the work because it's a different asset, because a different technology, is the safety profile. The safety profile of intismeran is similar as an infectious disease vaccine. So think if you have a cancer and you're diagnosed by cancer or a loved one, and you are told you can try this drug, and it's like you might spike a fever tonight, or you might have chills tonight, and that's it.
If you have cancer and that's a side effect profile, you're like, Yeah, sign me in. We've said it several times, but the phase III intismeran plus KEYTRUDA melanoma study was the fastest study, phase III of KEYTRUDA, that Merck ever enrolled. I cannot say we did it better than Merck. It's not true. Merck ran the study.
They are pretty good at running clinical study in oncology last time I checked. This was the fastest study ever enrolled in a phase III. You ask the investigators why, and I've done that myself, and they tell you, Stéphane Bancel, it was an easy call. We had very exciting data in phase II, a 0.5 hazard ratio, one in two people benefiting, and the safety profile is like our vaccine.
Every patient I told that they have one in two chance that this is better than what I'm going to give them anyway, KEYTRUDA, and there's no downside on safety, and they're like, Sign me in. Which is why the PIs sometimes will tell me, I signed up everybody because I recommend it to everybody, and when I explain it to them, they all signed up. That's I think the interesting thing about intismeran is its safety profile.
It's unlike anything I think we have seen as a field in oncology.
Absolutely. You're usually in some ways trying to-
It's a trade-off.
track and match.
It's always a trade-off.
Those eight.
When you see sometime, a lot of time, you add two checkpoints together, and then you have a lot of tox.
Mm-hmm. Absolutely. The core of your business today, though, is still very vaccines.
Yeah.
Very COVID oriented, we've obviously been through a lot of evolutions.
of the COVID market. It seems like we're getting to perhaps a more stable level, predictable level.
of what does COVID vaccine utilization look like going forward?
Yeah.
Perhaps you can make a comment on that, but also norovirus.
This is different to COVID in so many ways as you think about this opportunity. Can you talk about the market opportunity here, and what will it take for this to become a really meaningful contributor to Moderna's revenue profile?
Sure. Again, you have, again, lot of good question, Courtney, I want try to unpack everything. Yes. Today, we're an infectious disease vaccine business. We have now four vaccine approved around the world. We should get flu this year.
If norovirus phase III is positive later this year, we'll file it very quickly. It could be approved next year. If that happens, we will be, I believe, we're the largest respiratory portfolio of any company, which will be very helpful to us as we negotiate with customers. In the U.S., most people don't realize we sell directly to pharmacies, whether it's a CVS, a Walmart, local pharmacies, and there's no PBM in the middle, so we are direct in the B2B setting, and be able to bundle product is really important. For example, one of the reason we have struggled with RSV, in my opinion, is two thing. One has been the market has decreased in size tremendously, because of a lack of recommendation by ACIP for boosting people.
Basically, you got the first class of people who all went in, it was a great first year for GSK and Pfizer.
Year two and year three, we've no vaccination recommendation yet. I believe there will be vaccination recommendation because if you look at the epidemiology, natural infection, after a couple years, you start to be reinfected and sick again. What we've done over the last couple years is a few things. First, we've become much better at predicting the business because it was tough to model coming down from a pandemic. If you think about Moderna, we went from two big rollercoaster. Bless you. One was first going up from zero to $18 billion of sales in a year. I don't think a company has ever done that before. That's at least what Fortune had said at the time. We went down from $18 billion to around $1.9 billion last year.
That's down 90%. No pharma company has ever done that because when they lose patent expiry on one product, it's one product. This was the company.
We did everything we had to do in terms of resizing manufacturing, resetting the company, and so on. I think now we're in a very good place. We still have work to do on costs and opportunities, but I think we're in a very good place. What is interesting for me is last year we were in our guidance of.
for the first time, because the year before we were not. This year we're predicting growth.
If you look at how good our Q1 was, based on our outside the U.S. strategy, especially U.K. was a big driver in Q1. Even if we are flat in Q2, Q3, Q4, you will have sales growth compared to last year.
as you can run the math. I'm quite excited about where we're going. I think we stabilize, we're going back this year in sales growth, and we've a very diversified product portfolio now.
Four products approved, maybe five by Christmas, maybe six by next year, and a great geographic diversification.
which is of course very helpful. I'm very excited about norovirus because norovirus has no vaccine on the market. Everybody in the room, I'm sure knows of norovirus from personal experience. It's not a fun experience to go through.
I know doctors that have told me, When we have a norovirus vaccine, I'm sure going to get it every year, even though I don't get flu every year, and I know I should. I'll get norovirus every year, because I would rather have a flu on a weekend than having norovirus.
From doctors working in hospitals, protecting all of us. If you think about norovirus, the market is clearly people at high risk. Age, obviously, comorbidity, the usual, cancer, cardiac comorbidity, autoimmune disease, allergy, and other things that we know those infections are very bad for. The other one that's going to be interesting is healthcare workers. Like I just mentioned, the doctor I spoke to, I've heard it from nurses, nursing home with the elderly, and then teachers. Same, go talk to kindergarten teachers or lower school teachers, and they'll tell you, Yeah, I would like a norovirus vaccine. They might not get the flu shot or a COVID shot, but they'll take the norovirus vaccine because they get norovirus regularly because of just all the little kids they take care of every year, and that's how we parents get norovirus going back from school.
I think norovirus is going to be quite an interesting opportunity from a market size standpoint. I think it's going to be a very interesting opportunity for product bundling with customers.
If you look at the margin the retailers are making on vaccines, again, for people who don't know, they make profit from a discount they get from us and our competitors compared to list price. They get reimbursed by the payers, they also get an administration fee. When you add those two numbers, it's actually a very large number. If you do the math, you need to sell a lot of Kit Kat to have the same EBIT margin as selling one vaccine, which is why you get texts from CVS and Walmart all the time to get more vaccines and so on. That's why. When you get to a pharmacy, pharmacists will ask you, Oh, you're here for flu. Do you want COVID? Are you here for COVID? Do you want RSV? You want SHINGRIX? They get cash bonus, the pharmacist.
I think the ability to bundle products-
is really important for us in term of share. If you think about the growth of our customers, we're going to launch a new flu product. We are the only one with combo approved outside the U.S., but phase III data in the U.S. norovirus. We're going to be at a stage that was really handicapping us before, where we didn't have a lot of unique products. Moderna is literally in 2026, 2027, 2028, going to flip to a company that has not one, but several product that you're going to want for your growth-
that I'm going to be the only one able to provide to you. It's going to be quite interesting what happens to us in the next couple years. The good news for us from a P&L standpoint is I don't need to add one sales rep, so think about the cost leverage, because I still have the same team going to CVS and the same team going to Walmart. I don't need to add one sales rep in the U.S., and the manufacturing is all gravy because it's volume. I have a platform company. I don't need to spend $1 of CapEx to launch norovirus or the combo. It's the same manufacturing equipment, the same people in the same factory. What we're going to see, and by the way, the cost of goods of infectious disease, sorry, the R&D cost of infectious disease is going to come down.
because those phase III studies sunset. I'm going to have a very interesting P&L leverage at every line. Cost of goods, R&D, and SG&A.
Fantastic.
I love the vaccine business. By the way, I love the tail as well of a vaccine business. Those products last forever. I mean, you know, you used to work for Merck.
Yes, exactly.
Those products are on the market forever.
Some of them-
The barrier to entry is a billion-dollar phase III.
Absolutely, some of those products can grow for 40 years.
Correct
um, rather than-
Plus as we update them all the time.
The COVID this year is different from the COVID last year and the COVID two years ago.
There's also that aspect to it. I love infectious disease. I know not everybody does. I love it.
You touched on a few things that sparked my interest, and I do want to get to kind of some of the cost-cutting.
you've been going through and where there still is areas to cut. I think you made some comments about not needing to add to generate some of this revenue, which I think is interesting as well. There's also a question that's come in from the audience that builds on some of the vaccines topic. I'll hit that first, which is, we've seen an announcement on Tuesday this week.
that Eli Lilly has become a vaccines company.
again. three acquisitions
in one morning. Can you perhaps tell us, how does this sit with your strategy? How do you see them now as a competitive set? I think there was actually some commentary from Lilly that actually echoes some of the things that Moderna has commented on in terms of EBV.
and the potential for sequelae and long-term health impacts. It would be great to hear how are you internalizing kind of this new news?
I think it's first, great news for the world because we need more vaccines. There are more than 100 virus that hurts human. There's vaccines for less than 30. I think having a couple serious vaccine player is a good thing for humanity, number one. Number two, I think for everybody was running away from vaccine two years ago, and we stuck to it because we believe there's a medical need. If you add just the death of COVID, RSV, flu, and norovirus, it stacks in the top five deaths in this country.
Will you take a vaccine to not have cancer? Will you take a vaccine to not have a heart attack? Yes and yes, I'll take both.
Which is why every year, you know, I take my flu shot, my COVID shot. I got my first RSV shot because I'm more than 50 in the fall. Those are like table stake. You want to go for sport, sleep and run and eat healthy, and you don't take a vaccine. It's like, really? Okay. Then, as you said, I mean, EBV, we believe so deeply about EBV. We believe EBV is one of the nastiest virus that hurts human. Of course, we know it through mononucleosis, but it's most probably the number one cause of MS.
Yeah.
Many cancers are caused by EBV. A virus that stays in your body forever, it is not a good thing. It is not a good thing. Look at the new data coming on SHINGRIX with VZV on potentially slowing down dementia. It's only observational now. GSK is running a study, prospective study. It's not surprising. That virus stays in your brain. It inflames your brain. It's a piece of DNA. Those are DNA virus. It's a piece of DNA in your human cells. What does it do? It creates inflammation. If anybody likes inflammation, I don't think you.
you're thinking hard about biology. The world needs an EBV vaccine. We have one in phase II right now. Actually, we have two, sorry. One in phase II as a prophylactic vaccine to prevent mononucleosis, and hopefully, I would like to prove one day prevent MS.
Think about the medical impact of that. We have an EBV vaccine as a treatment for MS patients. It's in the clinic right now in MS patients as we speak, and it's trying to manage the viral load of EBV, so the quantity of copy of EBV virus, like people have managed beautifully HIV. If you think about HIV-
You try to manage the copies of the viral load of HIV to keep it undetectable so the virus doesn't do damage in your body. Same idea with EBV.
I'm delighted that Lilly is going after EBV. I think Lilly coming at this moment after what has happened over the last couple of years around vaccine hesitancy post-COVID, and what has happened around, you know, some vaccine, you know, misinformation.
As the biggest pharma company in the world that is saying, I want to deploy capital, not in one, not in two, but in three companies. I want to build a vaccine business. Like, hallelujah.
Fantastic.
I love it.
Yeah. Right.
I'm rooting for Lilly and for the patients they're going to protect.
Absolutely. Absolutely. You know, prevention is one of the best interventions that we can have.
Yeah.
Perhaps diving into that kind of costs and scaling-
side of the equation. I think in 2025, you took something like $2.2 billion of costs out of your business year-over-year. You're tracking to, I think, $4.2 billion in cash costs for this year.
That's the guidance.
Right.
Right, that we announced. That's the guidance.
kind of laid out a target for that even being lower-
by 2027, somewhere, I think, $3.5-$3.9-
Yeah
is the goal. Can you talk a little bit about where have you taken out cash costs? Where are you planning to?
to take out cash costs? Most importantly, where are you cautious?
Yes
when it comes to-
Yes
risking cutting too close to the bone?
Well, okay.
Ensuring that that revenue continues.
Yeah, good asking a lot of questions in one, Courtney. I'm going to, again, take them in pieces. First, we came from $9 billion of cash cost.
Last year, we had $4.3. The first wave was really cutting down manufacturing because we needed less volume. We had built a very large partner networks for filling, and we stopped most of them, but Rovi in Spain now. We had a big partnership with Lonza, which did roughly half of the drug substance of the mRNA and the lipid product during the pandemic. We also ended up the partnership with a big write-off for all those costs and the raw material that we had purchased because we didn't know the shape of a curve, and we wanted to make sure we could provide vaccine to protect people. It went down to prioritization of our portfolio R&D, so we cut in R&D.
What also helped us is some of infectious disease studies tailed off because a lot of people don't appreciate that for regulatory reason, all phase III study in infectious disease field lasts longer than the end of your phase III. For safety reason, and it's a good thing for the world, we have to monitor people in the studies several years after launch. When those costs start to tail off, of course, it helps you in terms of R&D cost. We took costs across the business. We restructured the whole business to right-size it for the work which was a streamlined portfolio. We worked a lot with suppliers on cost because during the pandemic, as you can imagine, even we had no time, we didn't go every time through an RFP process because an RFP process takes time.
Sometimes suppliers will tell you, Yeah, if you want this, you need to sign now, and it's signed now for three years, or you don't get product because I'm going to go to sell to AstraZeneca, my vials, or to Lilly or to somebody else. We knew as we went into the pandemic that there's a lot of work we were to unwind that cost structure that made no sense, but was the pandemic price of doing business. Nobody of the suppliers, I'm not throwing too many rocks at the suppliers, just a few. Nobody, if you are running a supplier company, whether you make vials or anything else, you had in your business plan for 2020 that there was going to be a pandemic.
Of course, you had no manufacturing infrastructure, you did not have people, you own the raw material to make those products. It was really, That's what I can do," and the capitalist system, right? We restructured a lot of that cost, and we're still doing it. I know because I was just with the procurement team a couple of days ago, and they're looking at the work they are doing for 2026 savings into 2027 for contracts that are expiring in 2026. We still have contracts that we have work to do to improve the cost structure. I'd say cost across the whole P&L, whether it's material cost or with some consultant or other things. Cost across the whole P&L. The R&D cost, as we talked about, as our RSV comes out, this will help as well.
The AI impact is going to be very profound. Because like most companies, we still have a lot of what we call AI gaps, meaning there's data in emails and in PowerPoint and in Smartsheet and in an ERP system. Sometimes it's not the same data. Sometimes humans have to check out what's true. It happened in every company I've worked for. It's happening at Moderna. It's even more acute at Moderna because we built the company so fast. If you think about Moderna in 2020, January 2020, we had no late-stage clinical capability, zero. We had zero commercial employees, zero. What we had to build in 2020 is all those function. We build them a lot in stacks parallel to each other. Why?
I hired the head of safety of Merck, and I said, Jose, I need a safety system in three months. What he did, he just threw warm bodies, great expertise. We went to steal to the industry, hire those people. We paid some suppliers for some IT system we needed that we didn't have, like IQVIA and other companies. All that now we are using actually AI to go and to figure out, if you could start everything from scratch and do it with today's technology, not even last year technology or two years ago technology, how would you design the business process? We're doing it business process by business process. I'm personally sponsoring several business processes in the company, and we have business processes across the enterprise and some business process within functions.
For example, Jamey in finance or CFO, he's working with his finance team on things within finance. How can they use technology to streamline how they do work in finance? I'm also challenging a lot of people about how we organize across functions. We're moving things also between functions, consolidating work. There's still work to be done. I would not be surprised if there's still a couple years of work to be done. I'm not saying you're going to see everything in cost because I'm going to still reinvest some of it into product or science that you will not see at the P&L level. This work reinvention that we're doing right now is, of course, massively enabled by AI and by vibe coding, obviously. It's going to impact every business process of the company.
I cannot think of a business process, and if you think of one, please tell me, that is not going to be impacted by our ability to reinvent it with much more digital capabilities. Not only AI, I mean, software that we can write ourself with an LLM through vibe coding, that we can plug to our system like SAP and so on, and then have, if we need a layer of LLM on top of it, to use AI on that data, but to do a lot of things, even in accounting. I think we can improve accounting tremendously. I think we can improve FP&A tremendously. I think we can improve purchasing. There's not one business process I cannot think of or I've not spent time with a team that I think we can improve in the next two or three years.
Because I want to pace things, because if I do everything at the same time, we're not going to get a good outcome. We need a good outcome because we are launching so many products, and we have so many good things happening at the company. We're just doing waves of business process, and we have very senior people sponsoring them, including me, as I said, for some of them, to make sure that the work gets done and that we really challenge more junior people about what really drives value. Like in any company, I still have people doing work that is interesting that doesn't drive value.
Yeah.
I'm using that also as a moment to say, okay, let's rethink about everything to build Moderna for tomorrow, because we have a platform company-
in terms of the science, the manufacturing. I don't think yet I have platformized the company in all the functions of a company. That's what we are doing now.
Fantastic. To that point, just continuing this AI part of the conversation-
Sure
You're obviously embedding this way of working.
Yeah
Perhaps this operational lens to deployment of AI. How can we do things more efficiently? There's also a lot of talk around drug discovery.
Can we crack new problems in biology?
Yeah.
Can we design targets or interventions that are-
much more likely to succeed in phase II and phase III?
Yes.
What's your kind of philosophy or thesis when it comes to the potential for the industry at large, but especially for Moderna when it comes to-
Yeah. I want to talk about Moderna because I think the industry will depend how quickly people adopt it, how quickly are they able to change how they do business.
Great.
I cannot talk about the industry, but for themselves. I can only talk about us. I'm extremely excited about AI for science.
I want to be realistic. Investors won't see it next year in the sales.
Yeah.
Right?
Because we're in the business, we need to develop product and take them to clinic.
Yeah.
For context, we have done machine learning in biology since 2016, 2017. This is not new to Moderna. We have actually invented enzymes, which are proteins, enzymes that we use in our factory, in our reactors that came in silico.
When this happened the first time, I remember vividly the team who put it. They all work in a computer, come up with an enzyme that we have not thought about that's going to be better than the one that exists in nature. Somebody was crazy enough to make it with their bare hands and test it. It was doing exactly as advertised by the computer. It was optimizing the feature that the scientists had coded in the machine learning system with, at the time, data scientists, because my scientists had no ability to write the machine learning algorithm. They had achieved what they wanted. They had improved an enzyme that exists in nature to there-
by basically, in my own view, accelerating time.
By just mutating amino acid on the enzyme, solving for the chemistry they were looking for while preserving a part of the enzyme that was doing the job they need to evolving another part of the enzyme that was doing something they didn't want happening in the reactors. That was a huge aha moment for me because trained as an engineer, I've heard about AI since I was in college, which is a long time ago. It was the type of thing that will never happen, always later. That moment was a huge aha moment for me. I was lucky it happened before COVID, because at the time, I set up an AI academy within Moderna.
I started training everybody on AI. Half my team thought I was nuts. Of course, COVID was a bit of a distraction to us. Of course, OpenAI happened with ChatGPT. When ChatGPT came out, I read about it a little bit, but a good friend of mine who is a software guy told me early December, he's like, You have to try this thing. I'm like, Really? He's like, Oh, yeah. You have to try this thing. At Christmas, calmly, with a bit of time, I went hiding somewhere with a computer for a couple hours, and my brain exploded.
I was doing science and then biology and then physics and then engineering and then French poetry and then Shakespeare, and I'm like, This is insane. Again, I think the story of Moderna, where I lived the S curve of mRNA technology from day one. The piece that shocked me, and it was a huge lesson for me because I've read in business book, and when I went to business school, I read about S curves.
I never lived one because when I was at Lilly or at bioMérieux, I was on the top of the S curve. 3% improvement per year was great. I've never lived the steep curve myself. I read about it in books. I was at Moderna, my scientists would come and say, We have a 5x improvement. I'm like, What? You have a what? A 5x improvement? They say, Yeah. I say, You already show me the data. Go redo the experiment. They come back, 5x. I say, Okay. Then two weeks after, another team will come, Oh, we have a 2x improvement, but that's uncorrelated to the 5x. Say, No, it's not. Say, That's 10x if I put those two piece of science together. I say, Uh-huh, it's 10x. This I'd never seen before.
The piece for me about AI, now we know because we have a bit of a few years behind us, if you talk to people, the hyperscalers and so on, or people at OpenAI or Anthropic or DeepMind, they'll tell you, as you've read, the doubling of compute every six month. Well, I can do math. You guys can do math, right? You can compound on your computer, right? I had that moment that December when ChatGPT came out, say, Oh, geez, this is going to be a big deal. Not necessarily now, but this is the worst version I will ever see in my lifetime.
It has been true every time I've got a new version. When we did that, I reached out to Sam Altman and said, We need your thing, but I cannot teach everybody else with my data. We did something, which basically we got a license for a product on AWS, but only for Moderna use.
We helped them develop ChatGPT Enterprise, which is now the product we use for all the employees of the company. In science now, I think it's going to change two big things for us. One is let's talk about the field. I think we are going to accelerate the pace of learning about biology. If you think about immunotherapy, that's changed the life of so many, created so much value. If you think about what we do in tisagenlecleucel and so many of the biology insights we have, we're accelerating as a field the pace of biology insights. We can now do, because biology is very complex systems with negative feedback loop in very complex biology. If you think about the biology we have done as humans so far in the last 50+ years is one gene-
one target. Most of the biology in our bodies doesn't work like that.
Which is why I'm excited about mRNA-2808, one of our products in the clinic, multiple myeloma, four mRNA in a drug. Going after three targets in one shot. That's the future of medicine. You cannot do that with recombinant, you can do that with mRNA. We've done already mRNA with 11 mRNA in monkeys.
I think the biology tailwind we're all going to get is going to be amazing. Those are going to be able to catch it, which will go back to the comment I made about the industry. The companies are going to be able to catch it and change how they work.
That the workforce will change how they work and be humble to sometimes say, I need to stop this job.
I need to structure a job differently. I need a different skill. I need to upgrade people. It's going to be a very painful process to adapt the workforce. It's a bit like when you work into a factory, when you went from people doing everything by hand, then having machines 100 years ago. It was a huge change. It's a huge change management. People who are going to have the courage to do it and the skills to do it are going to benefit tremendously. The second piece that's unique to Moderna is mRNA technology know-how.
What we're building now is a huge flywheel where AI-generated hypothesis on mRNA technology, I'm not talking biology, mRNA technology, build the mRNA, build the lipid. We can do 1,000 lipid at the same time. 1,000 lipid at the same time will take me several years before. We get that in one run. Do the in-vitro testing, go in animals, go in humans, feedback, learn, and keep going. I would argue there's no company in the world or public institution that has as much human data as Moderna. Same thing for monkeys. We're not scared at Moderna to go run 100 monkey experiment when we have to, with six or seven arms in the study to learn.
We're obsessed about learning about mRNA. That's one thing that has been since day one. We believe this is a very long game over several decades, and want to be the company who know the most about mRNA on the planet.
We want to use the data we have now in AI to learn even faster.
Yeah.
I want nobody to be able to catch up with us.
Yeah.
to bring that accelerated learning to patients through new medicine.
My goal is to be able to combine those two things. We're developing internal AI machine learning on mRNA, partnering with the best on the biology.
whether it's in Anthropic or on OpenAI, or DeepMind, Google, the best biology and access to all the platforms. I'm not wedded to any platform on biology. I want the best platform. Even the platform are changing each other. We have already processes in the company where you ask the three system the same questions.
You have them debate the same question in front of your eyes. It's fascinating. About very complex biology, and sometimes scientists realize, Huh, I didn't think about it.
A bit like you remember with AlphaGo, when AlphaGo came-
People are like, professional AlphaGo told you the computer made a move.
that no human had ever recorded. We're starting to see this in biology. We're starting to see biology insights that no human scientist that we're aware of had published on or documented.
Combining those two things to design drugs.
Yeah.
The scientists of Moderna, starting today, because I have pieces of those two vectors, I don't have the full vectors yet.
Getting those two vectors to be able to design new medicine is going to be fascinating. Of course, we have to take them to a clinic. The piece where I think the development process will be accelerated.
is like all my peers from GLP tox to filing an IND to phase I, phase II, phase III data, BLA, it's all a business process.
Sorry. It's all hundreds of business processes.
All of them you can apply AI vibe coding to shrink and shrink and shrink.
The ultimate thing, which most probably will take years, is can you start doing some clinical trial simulation in silico?
I don't think it's for tomorrow. We need FDA stability, as you described, and so on.
Yeah.
Given there's a finite number of cells in the human body, given we have 20,000 genes, given we have 40,000 metabolites, it's a very big math problem, but it's not an infinite math problem. Which is why you see us partnering with IBM in quantum two or three years ago.
Quantum is going to happen. I don't know precisely if it's happening in 2029, in 2030, in 2031. It doesn't change anything for our trajectory. When it starts working, when I can buy compute at IBM or Google or Amazon on quantum, I want my team to know about quantum, to understand quantum, to have already thought about how do we use quantum.
That's what we are doing. We've already had several problems that the team solved with quantum that we couldn't have solved before.
Otherwise.
That has already happened.
Yeah.
I want to be ready for when we need to run with quantum.
I don't want to discover, oh, we need to do quantum now.
It's not how we think about technology. I told you, AI 2018.
Yeah
AI academy at Moderna for all the employees. 2016, first machine learning. I'll do the same thing on quantum because I cannot even understand, literally my brain cannot understand what quantum and AI will do together.
Yeah.
Okay. I can put on compute quantum AI. I want to be able to solve very big science problem. I think one day we'll do clinical trial, not necessarily the full thing, but we'll do modeling of a clinical trial in silico.
And-
Much better prediction.
Yeah. You can predict all the biology. It's all science and math.
What happen in our body is not black magic. A lot of things we don't happen, it's not black magic.
Yeah.
It's chemical interaction. It's molecule binding to each other. It's all physics and chemistry.
Yeah.
Sometimes it's quantum physics. We've done quantum physics at Moderna.
to solve some problems. Sometimes you have to go very scientifically with it. It's all solvable. You just need enough compute.
Yeah.
That's why I'm very excited.
It's a very exciting future.
Yeah
Kind of particularly when you sit here and think, and I think others have said, maybe it's 1%, maybe it's 5% of biology that we understand today.
Yeah.
The amount of potential unlocks that exist in the future.
Yeah. Maybe just a side note.
is fascinating.
because it's public information. I'm on the board of Generate:Biomedicines. Generate:Biomedicines has a phase III drug that came out of a computer. The team designed it with some features they wanted on half-life. The computer designed the antibodies. It went into the clinic. It's in phase III now.
It's working in patients in phase II as advertised.
It's in phase III now. The notion that can we make drug using AI, I think is a question already answered.
Yeah.
The question is what type of innovative drug that you cannot even think about today you can make with that? We had a town hall after Q&A call, and I told the team because I spent a few days deep in science before that, as Lavina was preparing the Q call and everything, and I told them, "I'm convinced we have not invented our best drug yet.
I'm deeply convinced. I believe intismeran should be amazing for patients.
Mm-hmm. Mm-hmm.
PA, already the patient stories we're getting from parents for rare disease are unbelievable. The Christmas stories and so on, that they've never had a Christmas story like this.
All the hospitalization and deaths we're preventing with vaccines. I'm convinced, in every cell in my body, we have not yet invented our best medicine.
That's an amazing place to end this conversation. Perfect timing.
Yeah
kind of path towards innovation for the industry as a whole, but also kind of Moderna's journey within that industry. Thank you so much, Stéphane.