The AI-bio claim is real news, but thinly sourced
The most consequential AI claim in this dossier is also the least documented. Radical Numerics says AI-designed viruses have treated the startup's first patient, with an experimental treatment aimed at drug-resistant E. coli.1 That is the speaker's own statement, relayed in a technology newsletter. The dossier holds no trial registration, no outcome data and no peer-reviewed result. It marks the move from design to first dosing. It is not evidence the approach works.
Source quality matters here. The outlet that carried the claim, MIT Technology Review, has a measured reliability of 6% across 16 claims we checked against their sources, meaning only about one in sixteen held up.1 That score covers the outlet's newsletter content as a whole. It does not say this claim is false. It does say you should treat it as unconfirmed. The same newsletter family carried an unverified rejuvenation claim from Generation Lab, which says its treatment "blocks the systemic spread of aging in the bloodstream, reawakens the body's own repair mechanism and restores health and youth to multiple tissues."2 That is marketing language from the company. It is not a finding.
The topic brief also mentions an AI agent running gene-therapy injections. Our dossier contains no sourced detail on that, so we do not describe it.
The backdrop: money got more expensive in the same fortnight
Capital for early-stage AI-bio is tightening. On October 30, 2026, a $1 trillion tech selloff deepened fears of an AI bubble, as big AI bills and weak earnings spooked investors. The Nasdaq 100 had its worst day since June.3 The next day OpenAI cut its funding target by $10 billion.4 On November 1, Binance users withdrew $3.6 billion when the exchange reopened after a hack.5 The last item is crypto, not biotech, but it belongs to the same risk-off mood.
The link to biotech is simple. A company with no product revenue pays its bills from cash on hand and from new money. When investors pull back from AI-linked bets, the second source gets harder to tap. That puts cash runway and partnerships at the centre of the story.
What a funded biotech looks like: Incyte's verified cash
Incyte is not a start-up, but its SEC-checked numbers show how fast a balance sheet can swing. Its cash stood at $3.213 billion at the end of fiscal 2023 and fell to $1.688 billion at the end of fiscal 2024.6 It recovered to $3.098 billion at the end of fiscal 2025.6 The quarterly path is sharper. Cash was $987 million in Q2 2024 and $3.982 billion in Q2 2026, roughly four times higher.6 By Q3 2026 it was $3.276 billion.6 Put simply, a company with a long list of products can see its cash drop by half or more in a year and then rebuild it.
The same filings show costs rising. Cost of revenue was $104.5 million in Q1 2026 against $73.2 million in Q1 2025.6 It was $108.7 million in Q3 2026 against $99.0 million in Q3 2025.6 Capital spending is listed as $48.7 million in both fiscal 2024 and fiscal 2025.6 Identical figures in two years are unusual, and we flag that as worth checking against the filings. We would not read anything into it yet.
Incyte's footprint in the dossier is wide: Jakafi, Opzelura, pemigatinib, ruxolitinib cream and the pipeline compound INCB186748 are all listed as its products.7 The topic brief also names partnerships with BMS and Synnovation. Our dossier does not describe their terms, so we cannot size them.
The small caps: milestones are guided, cash is the test
The smaller companies are all promising 2026 and 2027 milestones. These are company guidance, not results.
Ernexa says it remains on schedule to submit its IND application for ERNA-101 in the third quarter of 2026, with a first-in-human Phase 1 trial expected in the fourth quarter of 2026.8 Prelude expects an IND submission for PRT13722 by mid-2026 and a Phase 1 study in ER-positive breast cancer in the second half of 2026.9 Prelude's mid-2026 date is already behind us, and the dossier shows no confirmation that it was met. Akari is further out: its CEO points to AKTX-101 advancing toward planned Phase 1 development in mid-2027.10 Akari argues that patients who no longer respond to antibody-drug conjugates are "a significant, growing unmet need" and an exclusive opportunity for its novel payload.10
Scribe is the nearest thing in this set to a technology story. It has listed on Nasdaq under SCTX, raised capital to fund the initial clinical development of STX-1150, STX-1200 and STX-1400, and expanded its collaboration with Sanofi.11 STX-1150 is an epigenetic silencing therapy for LDL cholesterol lowering, now in a first-in-human Phase 1 trial, with initial data expected in the first half of 2027.11 Scribe's pitch is that 50% to 70% of patients stop current LDL-lowering therapies within a year.11 It also cites reports that people with naturally occurring PCSK9 loss-of-function variants have 28% lower mean LDL-C and up to 88% lower coronary heart disease risk, and says STX-1150 aims to mimic that protection.11 Those are the company's statements of the scientific rationale, not trial results.
Vor is the one name with hard data. Its telitacicept Phase 3 TELIGAN trial in IgA nephropathy showed an annualized kidney-function decline of -0.6 against -5.7 mL/min/1.73 m²/year for placebo, a 5.1 difference (p<0.001), through two years.12 Risk of confirmed 30% and 40% kidney-function declines fell by 89% and 81%.12 These are company-reported final results, and none of it involves AI.
The large caps anchor the sector
Established revenue is what keeps the sector steady. BioArctic's royalty on Leqembi was SEK 179 million in Q2 2026, up about 10% on Q2 2025.13 Eisai reported Leqembi sales of JPY 29.3 billion for the quarter.13 BioArctic notes that Q2 2025 included a one-time Chinese stockpiling effect of around JPY 5.3 billion, and says that excluding it, royalty revenues grew about 43% year over year.13 Both growth figures are BioArctic's own. The 43% is the better measure of underlying demand, and the 10% is the headline. A reader should know which one they are being shown.
What to watch
Three tests will separate the claims from the results. First, whether Radical Numerics publishes anything beyond the first-dosing announcement. Second, Scribe's first Phase 1 data for STX-1150 in the first half of 2027, which will show whether epigenetic silencing works in people. Third, whether the funding squeeze reaches the smaller names. Incyte's cash swings show how much room a large company has. Pre-revenue companies have far less, and their guided milestones are only as good as the cash behind them.
Source reliability note: the NewsEOD wire items below carry a measured reliability of 54% across 6,692 checked claims. Company press releases are the company's own account. Treat guidance as intent, not outcome.

