Newsletter · 2026-09-20 · Ashutosh Agarwal

Thermo Fisher CEO Says AI Means More Lab Spending Not Less - Life-Science Tools Recovery Weekly - Week of September 20, 2026

The Life-Science Tools Recovery Weekly Recap for the week of September 20, 2026: a synthesis of the week's life-science-tools podcasts, headlined by Thermo Fisher CEO Mark Kasper breaking a six-week silence with concrete guidance (4% back-half growth, 7% organic by 2028) and the argument that AI drives more lab spending not less, Illumina's new CMO mapping a pivot beyond DNA into proteomics and the clinic, a former Illumina executive framing AI as data-starved, and a CDMO operator naming Sartorius, Repligen and Danaher's Cytiva on the single-use shift, while the bioprocessing book-to-bill stays missing.

Life-Science Tools Recovery Weekly

Week of September 20, 2026: Thermo Fisher CEO Says AI Means More Lab Spending Not Less

TL;DR

• After weeks of quiet, our two biggest names finally spoke, and both were on the front foot. Thermo Fisher's chairman and CEO sat for a rare on-camera interview and gave real numbers: 4% organic growth in the back half of 2026, with pharma and biotech (about 60% of revenue) growing faster, and a target to get back to 7% organic growth by 2028. His headline argument: artificial intelligence in drug discovery drives more lab spending on his products, not less. Separately, the Illumina executive who spent 31 years running the U.S. government's genome institute laid out where the sequencing giant is heading next: beyond DNA into proteins and other "omics," and out of the research lab into the clinic. This is the busiest, most tradable week for our sector in over a month, and it tilts the argument back toward the bulls.
• A credible former Illumina executive framed the single most important bull case for the whole sector: AI is starved for biology data. Omid Ostadan, now CEO of a private company called Cellanome, argued on a TD Cowen podcast that the bottleneck for AI in biology is not computing power, it's the lack of good data. Generating that data will be "one of the largest data-gathering endeavors ever undertaken by humanity," spanning years to decades. The TD Cowen analyst interviewing him said the quiet part out loud: that means "a lot more spending and wet lab work," exactly the instruments and consumables our companies sell.
• The one number that would settle the sector debate is still missing, but the read-through improved. For yet another week, no one who actually reports a bioprocessing "book-to-bill" (new orders versus shipments, the cleanest sign of whether the order book is turning) put a figure on the table. But a contract-manufacturing operator name-checked Sartorius, Repligen and Danaher's Cytiva directly, describing all three pouring money into the single-use and continuous-manufacturing equipment that is the future of the field. Encouraging on direction; still not an order read.

What's new

This was a loud week, the opposite of last week's near-silence, and, unusually, the loudest voices were operators at the companies we can actually trade. Ranked by what genuinely moves a portfolio.

1. Thermo Fisher's CEO breaks a six-week silence, with numbers and a clean AI argument.

The most valuable thing on the podcasts this week came straight from the top of our largest name. On Squawk on the Street (September 15), CNBC's David Faber sat down with Mark Kasper, chairman and CEO of Thermo Fisher Scientific, in a CNBC exclusive from Morgan Stanley's 2026 Global Healthcare Conference. This is an operator on the record: treat it as management talking about its own business (and, naturally, talking its own book), but it is the first real read from Thermo Fisher in weeks.

Start with the guidance, because it is concrete:

"When we think about this year in aggregate, we expect 4% growth and we expect to have 4% growth in the second half of the year. Pharma and biotech will be stronger than that. So that's our fastest growing end market. It's about 60% of our revenue. We've seen demand really pick up from the biotech companies."

Mark Kasper, CEO, Thermo Fisher

He also put a stake in the ground on the medium term: "You'll see us return to 7% organic growth in the not distant future," and when pressed, "we've said 2028 is our view. And it's progressing in the right direction." ("Organic growth" simply means sales growth from the existing business, stripping out acquisitions and currency swings.) He tied the recent pickup to what Faber called a "funding-to-revenue lag," the delay between biotech companies raising money and actually spending it on equipment. Recapping Kasper's fireside remarks, Faber noted that "the confidence from our customers and the emerging biotech companies is picking up," with funding starting "to improve as last year progressed" and stepping "up even further this year," a characterization Kasper did not push back on and built his guidance around.

The most quotable part was his answer to the AI-bear worry, the fear that if AI lets drugmakers do experiments on a computer, they'll buy fewer physical instruments and reagents. Kasper flipped it:

"AI is helping to kill ineffective medicines faster so that you don't spend money on things that don't work. And at the same point in time, shorten the time to market for those things that do work."

Mark Kasper

When Faber summed it up, "So, AI, more spending, not less spending. Is that what I'm hearing you say?", Kasper agreed: "That's our view in terms of, you know, how it will play out." His reasoning: better insight leads companies to validate more and chase more disease indications, so "that insight of what's likely to work actually gets more research being done." He also leaned on scale as a data advantage, "we have more data than anybody else in the life science tools industry, we're getting unique insights from that data," the same argument several operators have made all summer.

Two more nuggets that matter for the sector, because Thermo Fisher is the industry bellwether:

• On U.S. research funding (NIH): Faber noted the feared cuts "ended up being a lot less severe... given Congress restored lots of what the president wanted in cuts." Kasper's read: "we expect modest growth in NIH budgets over time. There's very strong congressional support for science... it's not going to be a strong growth market, but it should return to growth. And you saw us return to growth in the U.S. in the second quarter in academic." That is a notable softening of the funding gloom, more on why below.
• On China: Faber noted that on the Q2 call Kasper had said China "grew low single digits" but was "still not accretive to your organic growth." Asked whether he's seeing a return to growth, Kasper said "We are," adding it was "nice to return to growth in the second quarter" while cautioning he expects "slightly slower growth than the rest of the company." He was notably upbeat longer term: "when I came back from China in August, there's so much investment going on in pharma and biotech. I actually think that'll be a good growth market over time." As a former chair of the U.S.-China Business Council, he said he's spending more time on U.S.-China government-relations work.

He also flagged Thermo Fisher's leverage sits around 3.5x net debt to EBITDA after an acquisition closed in March, that he's comfortable it will come down as earnings grow, and that the company will stay active on deals in "a very fragmented industry."

Why it moves numbers: this is the clearest operator confirmation in weeks that biopharma demand is picking up, academic is turning in the U.S., and China is stabilizing, the exact three pillars the whole group needs. It is management's own framing, so weigh it accordingly, but it is the tradable signal of the week.

2. Illumina's new Chief Medical Officer maps the next chapter: beyond DNA, and into the clinic.

For the first time in a while we got a named Illumina executive on the record. On UnNatural Selection (September 15), Dr. Eric Green, Chief Medical Officer of Illumina, and previously 31 years at the U.S. National Human Genome Research Institute, 15 of them as its director, walked through where the sequencing leader is steering. This is operator commentary about strategy, not a demand or orders update, but it tells you how Illumina intends to keep growing now that the price of reading a genome has collapsed from "a billion dollars" to "a couple hundred dollars."

Green named two big shifts, and said Illumina was already ahead of him on both:

• Multi-omics, reading more than just DNA. "Illumina is really a multi-omics company now, as they should be." He pointed to two moves: the acquisition of SomaLogic, giving Illumina "a whole proteomics platform" (measuring proteins, not just genes), and a newly announced technology called 5BASE, "basically a technology for high throughput methylation analysis on the same platforms that you could do DNA sequencing" (methylation is a chemical mark on DNA that helps flag disease, cancer especially). The strategic point for investors: Illumina is widening what its machines measure, which widens the market it can sell into.
• End-to-end solutions, pushing sequencing out of specialist labs and into clinics. Green's vision: "at one end you squirt in some blood or some saliva or some DNA... off the other end comes a DNA report," with as few manual steps as possible, so hospitals and clinics worldwide, not just a handful of expert labs, can run it. He described Illumina "leaning into healthcare in a very serious way," shifting from a research-tools company toward clinical workflows.

He also flagged that Illumina "take[s] a lot of their revenue... and plow[s] it right back into R&D," and that the multi-omics wave depends on AI: "we won't be able to capitalize on multiomics without these advances in AI." Why it matters: this is the growth story Illumina is selling for the post-cheap-genome era: expand what you measure, and sell the whole solution into the clinic. It does not tell you what orders looked like this quarter, but it clarifies the direction and the recent deal logic (SomaLogic, 5BASE).

3. The cleanest bull case for the whole sector: AI is data-starved, and filling that gap means years of wet-lab spending.

On TD Cowen Insights (September 14), Omid Ostadan, CEO of Cellanome and a 15-year veteran of Illumina and its sequencing predecessor Solexa, made the argument that underpins the bull case for every tools company. (He now runs a private company, so treat him as an experienced, credible industry voice rather than an operator at a listed name.) The genome, he noted, now "costs less to sequence... than it does to go and buy an iPhone." But the bottleneck for AI in biology, he argued, is not computers or models, it's data:

"The rate-limiting step in activating AI in biology is the data, because the compute power is there... What's missing to a large extent is data."

Omid Ostadan, CEO, Cellanome

And generating that data, he said, "will end up being one of the largest data generation, data gathering endeavors ever undertaken by humanity... it's going to span certainly years, if not decades." The TD Cowen analyst interviewing him, Dan Brennan, connected it straight to the sector: "before computers begin to remove steps in the process, there needs to be a lot more spending and wet lab work," and asked whether that is "a good thing for the demand trends for a lot of the life science tools industry." That is the bull thesis in one sentence: the AI era doesn't shrink demand for instruments and reagents; it needs a mountain of new lab-generated data first. It echoes exactly what the Thermo Fisher CEO argued from the operator seat.

4. A bioprocessing operator names names, Sartorius, Repligen, Cytiva, but on technology, not orders.

We still did not get a bioprocessing book-to-bill from anyone who reports one. But we got the closest thing to a named-vendor read in weeks. On Off Script (September 15), Randall Vass, Executive Vice President of Process Design and Innovation at Just Evotech Biologics (JEB), a contract drug manufacturer (CDMO), described how biologics manufacturing is shifting toward smaller, more flexible "single-use" systems (disposable equipment you use once and throw away, rather than fixed stainless-steel tanks). Crucially, he named the suppliers benefiting:

"You've got the big vendors. Pick your favorite big vendors. Cytiva, Sartorius, Pall, you know, Repligen. And they're going to support you."

Randall Vass, EVP, Just Evotech Biologics

He singled out one directly: "Sartorius, in fact, has actually got almost an entire line of equipment that's designed to be fully continuous and they're still working on it, investing in it." He explained single-use gear actually lowers both cost and carbon footprint versus giant stainless tanks, and that protein A resin (used to purify antibodies) is "one of our most expensive components." He also flagged strong government interest, "there's a keen interest from the government to advance U.S. biomanufacturing," and JEB working with the FDA's emerging-technologies team. Why it matters: this is direct, on-the-record confirmation that Sartorius, Repligen and Danaher's Cytiva are the vendors of choice as the industry moves to single-use and continuous manufacturing. It is a technology-direction read, not an order read, but after weeks of nothing, it puts our bioprocessing names squarely in an operator's mouth.

5. Big money for late-stage drugs is available, but trial costs are exploding.

On The BioCentury Show (September 18), Nick Galakatos, Global Head of Life Sciences at Blackstone, described closing a $6.3 billion life-sciences fund (BLXS6) in March 2026, oversubscribed, at a hard cap, and 40% bigger than its predecessor, the largest private life-sciences fund of its kind. (This is investor commentary, clearly labeled, relevant because the health of biotech funding eventually flows into demand for lab tools.) His warning for the ecosystem: the cost of late-stage clinical trials has "doubled, tripled, quadrupled," and "it's not uncommon now to see a phase three program cost in excess of a billion dollars," on top of "pricing pressures, geopolitical pressures, and many other pressures that compress the amount of capital that the industry can put into development." The read for us is two-sided: serious capital is flowing to late-stage biotech (good for downstream demand over time), but rising trial costs and pricing pressure squeeze how much of each dollar reaches the lab bench.

The debate

Every week we steel-man both sides. This week the bull got the stronger and more concrete evidence, a near-mirror image of last week, when the bear had the better witness.

The bull (bioprocessing recovery / sequencing and multi-omics re-acceleration). The bull had its best week in over a month, and, for once, from operators, not just pundits. Thermo Fisher's CEO confirmed demand is picking up, guided to 4% growth in the back half with pharma/biotech faster, targeted 7% by 2028, said U.S. academic returned to growth in Q2, and sounded optimistic on China. The AI argument landed cleanly from two independent seats: Thermo Fisher's Kasper ("AI, more spending, not less") and former-Illumina executive Ostadan ("the rate-limiting step... is the data"), backed by a TD Cowen analyst spelling out "a lot more spending and wet lab work." Illumina, meanwhile, is visibly expanding its addressable market (proteomics via SomaLogic, methylation via 5BASE, plus a push into the clinic). And a CDMO operator confirmed Sartorius, Repligen and Cytiva are the go-to vendors as manufacturing moves to single-use and continuous systems. That is a lot of genuinely positive, named signal.

The bear (funding / trial-cost inflation / lumpy demand / no order read). The bear's case is quieter but not gone. First and most important: still no bioprocessing book-to-bill from any listed supplier, the JEB commentary names vendors investing in equipment, but says nothing about whether orders are actually turning. Second, Blackstone's Galakatos flagged that phase-three trial costs have risen two-to-four-fold to over $1 billion, with pricing and geopolitical pressures compressing industry capital, a slow squeeze on how much reaches the lab. Third, the most bullish voices this week are self-interested: a CEO and a company CMO talking their own books, plus a former-Illumina executive on a sell-side-hosted podcast. And there was no read at all on the head-to-head sequencing competition (Illumina versus Element, Ultima, MGI, PacBio, Oxford Nanopore), and nothing concrete on tariffs or currency.

Our read. The honest scorecard tilts bullish this week, a clear reversal from last week, and mostly on the strength of finally hearing from operators at our two biggest names. The demand narrative (biopharma picking up, academic turning, China stabilizing, AI = more spend) got real operator confirmation for the first time in weeks. But keep the caveats sharp: this is management framing, not independent data, and the single number that would end the whole sector debate, the bioprocessing order book, is still missing, now for many weeks running. Expect it to stay missing until Q3 earnings season begins in late October. The direction of travel improved a lot this week; the hard proof did not arrive.

Stocks in play

Every ticker in the coverage universe. This week Thermo Fisher and Illumina got direct operator commentary; Sartorius, Repligen and Danaher (via Cytiva) were named by a CDMO operator on technology (not orders); the rest got read-through only. Where a name got nothing specific, we say so and carry the standing debate forward.

Companies referenced this week that are not in the coverage universe: Cellanome (private; CEO Omid Ostadan), Just Evotech Biologics / JEB (private CDMO), SomaLogic (proteomics, acquired by Illumina), and Blackstone Life Sciences (Nick Galakatos) are not investable here directly, but they map the demand direction and the deal logic for the listed names above.

Read-throughs

• Biopharma capex / instruments, the week's best signal. Thermo Fisher, the sector bellwether, gave the first clear operator confirmation in weeks that demand is picking up: 4% growth guided for the back half, pharma/biotech (~60% of revenue) faster, and a "funding-to-revenue lag" now turning positive as biotech funding that improved last year and stepped up this year finally reaches the lab. Bullish read-through for the whole instruments-and-consumables group (Thermo Fisher, Danaher, Agilent, Bruker, Waters, Mettler-Toledo, Revvity, Bio-Rad).
• Academic / NIH funding, a notable softening. This is where the story changed most versus last week. Where Francis Collins described "unprecedented" NIH cuts seven days ago, the Thermo Fisher CEO this week said the cuts "ended up being a lot less severe... than had been anticipated," with Congress restoring much of the proposed reduction, and, importantly, that Thermo Fisher "returned to growth in the U.S. in the second quarter in academic," with "modest growth in NIH budgets over time" expected. Not a boom, but a real de-escalation of the funding fear that has hung over the group all summer. Relevant to every academic-exposed name (Illumina, 10x, PacBio, Bruker, Agilent, Bio-Rad, Avantor, Thermo Fisher).
• China, a first real operator read in a while. Last week China was "a complete blank." This week the Thermo Fisher CEO put a number and a view on it: China grew low-single-digits in Q2 (not yet adding to company growth), but after an August visit he sees "so much investment going on in pharma and biotech" and expects "a good growth market over time." One operator, one company, but after weeks of silence, it is the first genuine China demand read for a Western instrument maker. Relevant to all China-exposed names (Thermo Fisher, Danaher, Agilent, Bruker, Mettler-Toledo, Waters).
• Bioprocessing / single-use and continuous manufacturing. Still no book-to-bill from anyone who reports one, the multi-week gap continues. But a CDMO operator named Sartorius, Repligen and Danaher's Cytiva directly as the vendors investing in, and benefiting from, the shift to single-use (disposable) and continuous manufacturing, with government backing for U.S. biomanufacturing. Good for the long-run technology direction of the bioprocessing suppliers; not an order read. Net: direction reinforced, order book still dark.
• Sequencing / multi-omics. The competitive picture (Illumina vs Element / Ultima / MGI / PacBio / Oxford Nanopore) got no direct read. Instead the news was strategic and structural: Illumina expanding beyond DNA into proteins (SomaLogic) and methylation (5BASE) and into the clinic, and a strong AI-needs-data demand thesis (Ostadan, echoed by the Thermo Fisher CEO). Bullish for sequencing and single-cell/spatial demand over time (Illumina, PacBio, 10x); silent on who is winning the head-to-head instrument war.
• Biotech capital / M&A. Two-sided. Blackstone's $6.3 billion fund shows serious money is available for late-stage biotech, but trial costs up two-to-four-fold (now over $1 billion for a phase-three program) plus pricing and geopolitical pressures squeeze the capital that reaches the bench. Separately, on the same Thermo Fisher episode, Morgan Stanley's global co-head of M&A said 2026 healthcare deal volume has already matched all of 2025 in eight and a half months and does not see it slowing, supportive of continued consolidation among the tools and pharma names.
• Tariffs / reshoring / FX. On Molecule to Market (September 18), an industry consultant described duties and U.S. pricing policy pushing pharma to move manufacturing onshore, "there will be a regionalization of supply chains," with the U.S. "the country where the government has been most determined to make a change." A slow structural tailwind for equipment and consumables sold into U.S. plants, consistent with the reshoring thread we have tracked for weeks. No direct read on currency (FX) this week.
• Cell-and-gene-therapy (CGT) reagent demand. No direct operator read on CGT reagent volumes this week (relevant to Maravai). Carry forward as unconfirmed.

What changed vs last week

Last week (September 13) was the quietest tape in over a month, zero by-name operator commentary on any of our 15 companies, with the loudest voice a bearish one: Francis Collins on "unprecedented" NIH cuts. This week is a sharp reversal on both counts: the operators showed up, and the tone flipped bullish. Concretely:

1. Our two biggest names finally spoke, and both were constructive. Last week: total operator silence on all 15 tickers. This week: Thermo Fisher's CEO with hard guidance (4% in 2H, 7% by 2028, biopharma faster, demand picking up), and Illumina's CMO laying out the multi-omics and clinical strategy. That is the single biggest change, the tradable names went from dark to on the record.
2. The NIH-funding story softened, and the update came from an operator. Last week Francis Collins described cuts already landed and a demoralized workforce. This week the Thermo Fisher CEO said the cuts "ended up being a lot less severe... than had been anticipated," that Congress restored much of the proposed reduction, and that Thermo Fisher's U.S. academic business "returned to growth" in Q2. Same topic, materially less scary read, and this time from a company that actually sells into those labs. Both can be true, Collins on the projects already cut, Thermo Fisher on the aggregate turning, but the near-term demand read is clearly better than last week's.
3. China went from "a complete blank" to a real, if single, operator read. Last week we flagged zero read on Chinese demand for Western instruments. This week the Thermo Fisher CEO gave one: back to low-single-digit growth in Q2, and optimistic after an August visit. Still one voice, but a genuine data point where last week had none.
4. Bioprocessing color got more specific, named vendors, not just adjacent themes. Last week's texture came from CDMO operators (Siegfried, SK pharmteco) on reshoring and consumables intensity. This week a CDMO operator named Sartorius, Repligen and Cytiva directly as the single-use/continuous vendors of choice. A step up in specificity, though, again, on technology, not orders.

What did NOT change, and it is still the crux: there is still no bioprocessing book-to-bill from any listed supplier (Sartorius, Repligen, Danaher/Cytiva named this week, but on equipment, not orders); still no read on the head-to-head sequencing competition; still nothing on currency; and several names, Agilent, Revvity, Bio-Rad, Bruker, Waters, Mettler-Toledo, Avantor, Maravai, 10x, PacBio, got no operator commentary at all. The mood improved sharply this week and the funding fear eased, but the one number that would actually settle the sector debate, the bioprocessing order book, is still missing. Expect it to stay missing until Q3 earnings season begins in late October.