Forest Neurotech

Phase A — Understand the business

Lens 1 · Company Overview

What it is — and the category error to avoid. Forest Neurotech is a nonprofit Focused Research Organization (FRO) operating under Convergent Research (the FRO incubator founded 2021 by Adam Marblestone and Anastasia Gamick; ~10 FROs created to date). FROs are "startups for science" — fixed-term, mid-scale engineering pushes that build tools other scientists use, explicitly NOT companies pursuing a P&L or an exit. This is the single most important framing in the dossier: Forest is structurally untradeable. There is no cap table to buy into, no path to a ticker, no equity.

The product. Forest's flagship is Forest 1 — described by Forest and by funder ARIA as a minimally invasive, whole-brain computer interface that uses ultrasound (not electrodes) to both measure (functional ultrasound imaging, fUSI) and modulate (focused ultrasound, FUS) brain-wide activity. The hardware is a research-grade ultrasound scanner "smaller than a standard key fob," powered by Butterfly Network's Ultrasound-on-Chip™. It is designed to sit implanted flush with the dura beneath an acoustically transparent cranial window — minimally invasive relative to penetrating electrode arrays, but still a surgical implant (it is not a wearable-on-skin device, despite some loose press framing).

Customers / go-to-market. Forest's stated model is license the platform to academic and industry research organizations — and it has said explicitly: "Within the lifetime of our nonprofit research organization, we don't have plans to submit to the FDA". So the "customer" is the research community, not patients or payers. The nearest thing to a commercial counterparty is its supplier, Butterfly Network (see Lens 2), and its funder/validator, ARIA + the NHS (see Lens 5/10).

Contract structure / key terms. The economically concrete relationship is the Butterfly co-development agreement (Lens 2): a 5-year deal under which Forest pays Butterfly up to $20M (licensing + chip purchases + services + milestones), $3.5M already paid, plus per-unit revenue to Butterfly on every device sold. Note the direction of cash: Forest is a payer, not a revenue generator.

Bottom line (Lens 1). Forest is a philanthropically-funded research instrument whose job is to de-risk the ultrasound-BCI modality and prove it in humans. Its commercial value does not accrue to "Forest" — it accrues to whoever productizes the validated modality. As of Jan 2026, that party has a name: Merge Labs (Lens 9, Lens 11).

Lens 2 · Supply Chain (named stakeholders)

Map of the actual named entities along Forest's chain:

Chokepoints. (1) Butterfly is single-source for the sensing chip — the entire device pipeline depends on one NYSE-listed supplier with its own going-concern history. (2) The IP boundary with Merge Labs is undefined — if the valuable IP migrated to Merge, Forest is left holding a research mandate without the crown jewels. Names here, not generic: the chain is Caltech → (Butterfly silicon) → Forest → NHS, with Merge Labs as a fork off the founder node.

Lens 3 · Competitive Advantages (moats)

Where the edge is real:

• Modality differentiation. Ultrasound is genuinely differentiated from the entire electrode field (Neuralink threads, Synchron Stentrode, Precision's surface array, Paradromics). Two structural advantages: (a) whole-brain / deep reach — focused ultrasound can address deep structures from a single device, where "electricity does not travel very far through brain tissue" so electrode systems must be physically placed at each target; (b) less invasive — only an acoustically transparent cranial window is implanted, not penetrating arrays through the dura/cortex.
• Read and write in one modality. fUSI images activity and FUS modulates it — a combined sense-and-stimulate loop competitors split across modalities.
• Scientific provenance / talent moat. The Caltech Andersen + Shapiro lineage is a genuine credibility moat in a field where the science is the scarce input (Lens 9). Peer-reviewed proof points in Nature Neuroscience (2023) and Science Translational Medicine (2024).
• Butterfly chip access. Exclusive-feeling access to a manufacturable ultrasound-on-chip compresses years of hardware work.

Where the "moat" is weak or misframed:

• A nonprofit has no economic moat. Forest's mandate is to publish and license broadly, which is the opposite of a defensible commercial position. Any moat it builds, it is structurally inclined to give away.
• The physics ceiling is the bear case (Lens 13). fUSI measures hemodynamics (blood flow), not spikes — a 6-second peak latency, optimizable to ~2.4s but never millisecond. That is a modality-level limitation, not a Forest execution gap, and it caps the addressable use-cases.
• The founders/IP can leave — and did. A talent moat you don't own evaporates when the talent spins out a for-profit (Merge Labs). Bargaining power over its key supplier (Butterfly) is also weak: Forest is a payer of up to $20M, not a price-setter.

Bargaining power. Low over Butterfly (single-source, Forest pays). Moderate over the research-tool "market" (novel capability, few substitutes for deep+whole-brain access). None in the commercial sense — it's a nonprofit.

Lens 4 · Segments

segments.csv is empty (header row only). Forest has no revenue segments — it is a pre-revenue nonprofit. There is nothing to break out by product or geography in the operating sense. The only meaningful "segmentation" is by funding source and program, reconstructed web-only:

Geography: US-incorporated nonprofit; flagship clinical activity in the UK (NHS) — a deliberate regulatory-arbitrage choice (UK first-in-human is faster than a US IDE; reinforced by Forest's stated no-FDA posture). Trend: the funding base is shifting from US philanthropy (2024) to UK government-clinical (2025), while the commercial center of gravity has shifted entirely out of Forest into Merge Labs (2026).

---

Phase B — Measure performance

No financial tape exists. Forest has no earnings, no consensus, no public stock, no transcripts. Phase B is therefore re-pointed per the +private overlay to funding/valuation trajectory, traction, and milestone catalysts, with the brutal caveat that the company is pre-revenue and unaudited.

Lens 5 · Funding & Valuation Trajectory (overlay swap for "Earnings Result")

Forest's own trajectory:

• Oct 2023 — Butterfly 5-yr co-development agreement; Forest commits ≤$20M to Butterfly ($3.5M paid up front).
• Mar 2024 — $14M philanthropic grant, Schmidt-led (Eric Schmidt, James Fickel, Griffin Catalyst, Riley & Susan Bechtel Foundation).
• Jan 2025 — £6.5M ARIA award for the NHS Forest 1 first-in-human trial.
• No priced equity round, ever. As a nonprofit FRO there is no post-money/pre-money valuation — "valuation: n/a — nonprofit, not disclosed."

The valuation that actually matters is its successor's. Merge Labs (the Forest spinout) raised a ~$252M seed at an ~$850M valuation in Jan 2026, OpenAI writing the largest check. That $850M is the closest market-clearing price for "the Forest ultrasound-BCI thesis," but it is Merge's mark, not Forest's.

Burn / runway signal. Forest is consuming philanthropic + grant capital (~$22M+ disclosed inflows 2024–25) against a ≤$20M multi-year Butterfly obligation and a clinical trial. As an FRO it runs on a fixed-term budget, not perpetual fundraising — the risk is mandate completion, not insolvency. Unaudited per public sources.

Lens 6 · Founder Interviews / Public Sentiment (overlay swap for "Earnings Calls")

No earnings calls exist. Substitute: founder/CEO public commentary trend.

• CEO Sumner Norman has been heavily on the speaking circuit — EPFL Neuro-X seminar ("Toward an ultrasonic whole-brain-computer interface"), AI for Good Summit (ITU), the Foresight Institute podcast ("Neurotech for Mental Health"), SynBioBeta, Bionics Gamechangers, Core Memory podcast. Recurring message: ultrasound as the less-invasive, whole-brain alternative; mental-health (depression/anxiety) as the lead application; the science is real and peer-reviewed.
• Tone shift over time: through 2023–2025 the framing was mission-driven nonprofit science ("a tool to understand the brain," "no FDA plans"). In 2026 the energy and the commercial narrative migrated to Merge Labs — the same founders now attached to a $252M for-profit with a far more aggressive "thought-to-ChatGPT" framing. The thing they stopped saying: the purely nonprofit, no-commercial-exit framing. That is the single most telling sentiment shift in the file.

Lens 7 · Cap Table & Comparable Marks (overlay swap for "Comps")

Forest's "syndicate" is philanthropic, not a venture cap table: Eric Schmidt, James Fickel, Griffin Catalyst (Ken Griffin), Riley & Susan Bechtel Foundation, ARIA (UK gov). No tier-1 VC, no crossover fund — because there is no equity to own. The IPO-proximity tells the +private overlay looks for (Fidelity/T. Rowe/Coatue entries) are structurally absent and always will be for Forest itself.

The comp table that matters — the ultrasound-BCI thesis vs. the electrode field:

TAM frame: Grand View / industry estimates put the BCI market at $8–12B by 2030. Read this table the right way: the modality Forest pioneered is now being valued by the market at $850M (Merge) against a field leader at ~$9.6B (Neuralink) — i.e., ultrasound is the early-stage, highest-uncertainty entrant, priced ~11x below the leader.

Lens 8 · Catalysts / Events (overlay: funding & product events, not stock moves)

No stock, so the "what moves the value of this thesis" timeline:

• Oct 2023 — Butterfly co-dev deal announced (validated manufacturable hardware path).
• Nov 2023 — Caltech closed-loop ultrasonic BMI paper, Nature Neuroscience (science de-risk).
• Mar 2024 — $14M Schmidt-led grant (capital de-risk).
• 2024 — Science Translational Medicine fUSI-through-cranial-window-in-humans paper (human-feasibility de-risk).
• Jan 2025 — £6.5M ARIA award + NHS trial announced; press calls Forest 1 the "most advanced" BCI.
• ~Mar 2025 — NHS Forest 1 trial slated to commence (3.5-yr program, first ~8 months for regulatory approval) — the de-risking catalyst that matters most.
• Jan 2026 — Merge Labs emerges from stealth, $252M / $850M, OpenAI-led — the catalyst that revealed where the equity value went.

Pattern read: every Forest catalyst de-risks the modality (which benefits Merge and Butterfly), not Forest's own balance sheet — because Forest captures no equity upside. The market "reacts" by funding the for-profit successor, not the nonprofit.

---

Phase C — Judge people & books

Lens 9 · Management (founder archetype)

Sumner Norman — Co-founder & CEO. PhD in brain-machine interfaces / human-robot interaction; postdoctoral fellow in Richard Andersen's lab at Caltech, where he co-first-authored the foundational ultrasonic-BCI work (decoding movement intentions from fUSI in non-human primates). Co-founded Forest in 2023 on that research. Track record: primarily scientific/translational — genuine peer-reviewed breakthroughs, less a record of building/scaling an organization. Archetype: scientist-founder. Skin in the game: n/a in equity terms at Forest (nonprofit); now a co-founder of for-profit Merge Labs.

Tyson Aflalo — Co-founder. Caltech (Andersen lab) neuroscientist; co-founder of Forest and also a co-founder of Merge Labs.

Mikhail Shapiro — scientific anchor. Caltech professor; pioneer of acoustic biomolecules / ultrasound neuromodulation; advisor/collaborator to Forest and a named co-founder of Merge Labs. His public quote — "Using blood flow to image neural activity with functional ultrasound works really well" — is also the source of the bear case (it's blood flow, not spikes).

William Biederman — CTO. Cited as Forest's CTO; emphasizes stimulation+recording "with submillimeter accuracy".

Convergent Research / Adam Marblestone — the FRO parent's CEO; sets the nonprofit operating model.

Capital-allocation read. As a nonprofit, "capital allocation" = grant deployment toward a fixed scientific milestone (human Forest 1 data). No buybacks/M&A frame applies. The defining capital-allocation event is the Merge Labs spinout — value created inside a philanthropically-funded nonprofit became the seed for an $850M for-profit owned by (some of) the same founders + Altman + OpenAI. Whether that was a clean, arms-length licensing of Forest IP to Merge — or value leaving the nonprofit on favorable terms — is the governance question (Lens 10/13); it is "not been clarified publicly".

Red flags (people). No fraud signals. The flag is structural conflict: the same individuals run a donor-funded nonprofit and a for-profit pursuing the same modality, with OpenAI (Altman's other company) as lead investor. Donor capital de-risking a privately-owned for-profit is a real reputational/governance risk even if entirely above board.

Lens 10 · Forensic Red Flags + Regulatory

Accounting forensics: n/a — no audited financials exist. As a US nonprofit, Forest would file an IRS Form 990 (not reviewed this pass — flag as an open item; a 990 would show comp, grants, and the Merge-related transactions and is the single highest-value document to pull next). Revenue recognition / SBC / goodwill lenses don't apply to a pre-revenue nonprofit.

Regulatory findings:

• SEC (EDGAR EFTS — LR + AAER): No CIK; not an SEC registrant; zero findings possible — total_sec_findings: 0.
• Non-SEC enforcement (web search): Searched "Forest Neurotech (FTC OR DOJ OR FDA OR consent decree OR settlement OR fine OR penalty) enforcement." No material enforcement, litigation, fine, or consent-decree hits found across the press corpus.
• FDA: Forest has publicly stated it does not plan to submit to the FDA within the org's lifetime; its clinical activity runs under UK/NHS regulatory approval (the ARIA-funded trial's first ~8 months are dedicated to obtaining UK regulatory clearance). No FDA enforcement surface because there is no FDA submission.
• Item 3 (Legal Proceedings): n/a — no 10-K exists (private nonprofit).

Conclusion: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER → 0), web search (no hits), and confirmation of no FDA submission, as of 2026-06-23. The only genuine "forensic" question is the governance of the Forest→Merge value transfer (Lens 9/13), which is a transparency gap, not a known violation.

---

Phase D — Project & stress-test

Lens 11 · IPO-Readiness & Path-to-Tradeable (overlay swap for "Forward Projection")

No EPS projection — Forest is a pre-revenue nonprofit. No Brier forecast logged (per --watchlist rules and the SKILL's no-forecast-on-watchlist guidance). The +private Lens 11 asks: what unlocks a tradeable security, and when? For Forest the answer is stark.

Forest itself: path-to-tradeable ≈ none. As a nonprofit FRO with no equity, Forest will never IPO and cannot be bought as a security. It is absent from research/private-watch.json for exactly this reason — there is no IPO readiness to score. The correct entry in that ledger, if added, would be ipo_readiness: n/a — nonprofit; tradeable exposure is via Merge Labs (private) or Butterfly (BFLY, public).

The tradeable exposures the thesis actually creates:

1. Merge Labs (private, ~$850M, Jan 2026) — the real equity vehicle for the ultrasound-BCI thesis. Its IPO-readiness: very early (seed stage, no disclosed product/clinical timeline, modality still the primary uncertainty). Window to S-1: multi-year, gated on first-in-human ultrasound-BCI efficacy data. Access today only via late-stage private/secondary or an OpenAI-orbit allocation.
2. Butterfly Network (NYSE: BFLY) — the only liquid, today-tradeable proxy. Public, supplies the chip, earns ≤$20M + per-unit from Forest, and is the picks-and-shovels play on all ultrasound-BCI efforts (Forest and Merge both need ultrasound-on-chip). The cleanest way to express "ultrasound BCI is real" in a public portfolio. (BFLY's own fundamentals not analyzed this pass — flag for a dedicated BFLY dive.)

Milestones that unlock value (for Merge, and indirectly BFLY):

• NHS Forest 1 first-in-human safety + mood/motivation readout (3.5-yr trial from ~Mar 2025 → ~H2 2028) — proves the modality in humans.
• Any Merge Labs product/clinical disclosure (none yet as of Jan 2026).
• A second priced Merge round (would re-mark the thesis).

Lens 12 · Bull vs Bear

Bull case (the thesis at its strongest). Ultrasound is the right long-term modality and Forest's science proves it: whole-brain + deep reach + read-and-write in one device + dramatically less invasive than penetrating electrodes, with peer-reviewed primate and human-feasibility data and a manufacturable Butterfly chip. The market agrees enough to fund the successor (Merge) at $850M with OpenAI as lead. If fUSI's latency problem is engineered down and the NHS trial shows a real mood/motivation effect, ultrasound leapfrogs the surgical-electrode field on the dimension patients care about most — not needing brain surgery. For Forest's mission, the bull case is "it works, gets licensed broadly, and seeds a whole sub-industry" — which is arguably already happening.

Bear case (2–3 things that permanently impair the THESIS — Forest has no business to impair).

1. The physics ceiling is structural, not solvable by money. fUSI reads hemodynamics, peaking ~6s post-stimulus, optimizable to ~2.4s but never millisecond. High-bandwidth, real-time BCI control (the use-case driving Neuralink's $9.6B) may simply be out of reach for a blood-flow signal. If so, ultrasound is confined to slower mood/state monitoring + neuromodulation — a smaller prize than bulls imply.
2. Forest captures none of the upside. Even in the full bull case, a nonprofit does not convert success into shareholder value. The capture went to Merge Labs. Owning the thesis via Forest is impossible by construction.
3. Modality risk is still unpriced. Even Merge's own backers concede "the modality itself remains the primary uncertainty," and Merge committed Neuralink-early capital with no specified product or clinical timeline. An $850M seed on an unproven modality is a fragile mark.

Pre-mortem (18 months out, the thesis broke — what happened?). The NHS Forest 1 trial reads out null or ambiguous on mood/motivation; or it surfaces a safety/imaging-stability problem under the dura; meanwhile the hemodynamic-latency wall proves real in closed-loop human use, capping ultrasound to niche monitoring. Merge's $850M mark looks rich, its next round is flat or down, and the "ultrasound leapfrogs electrodes" narrative deflates while Synchron/Neuralink ship FDA-track products.

Are multiples too high? For Forest, n/a (no equity). For the thesis, Merge's $850M seed is the stretched number — Neuralink-stage capital on a pre-clinical-validation modality.

Contrarian view (what the market refuses to see). The market is treating "Forest Neurotech" as a BCI company to track for a position. It is not — it is a nonprofit research subroutine whose entire economic output was designed to be given away and has already been forked into Merge Labs. The contrarian, correct move: stop trying to own Forest; track it as a free, high-quality science-de-risking signal, and take the position (if any) in Butterfly today or Merge later.

Lens 13 · Devil's Advocate (short-seller)

Dismantling any "go long Forest / the ultrasound-BCI thesis" case:

• You literally cannot short or long Forest — there's no security. So the only tradeable target is Merge ($850M private) or Butterfly (BFLY public). The short case attaches to Merge's valuation and to BFLY as a Forest-revenue proxy.
• Revenue concentration (Butterfly): the entire Forest-derived revenue to Butterfly is a single ≤$20M, multi-year, milestone-gated relationship with a nonprofit that has said it won't commercialize via FDA. If the modality stalls or Merge in-sources its own transducer, even that thin revenue stream is at risk. Don't capitalize BFLY on Forest.
• The moat may be weaker than bulls think: the "moat" (Caltech talent + IP) already walked out into Merge. Forest is the cautionary tale — a talent moat you don't own.
• Most dangerous competitor bulls underestimate: not another ultrasound shop — it's the electrode incumbents shipping FDA-track clinical products (Synchron's 2026 pivotal trial; Neuralink's expanding human cohort). They could lock in the high-value indications (motor restoration, communication) while ultrasound is still proving feasibility, relegating ultrasound to slower mood/neuromodulation niches.
• Worst governance optics: donor capital (Schmidt, Griffin, Bechtel, UK taxpayers via ARIA) de-risking a modality that the same founders then commercialize in a privately-held, OpenAI-funded for-profit (Merge), with the IP/personnel split "not clarified publicly." A skeptic prices that as either reputational risk or a sign the nonprofit was an incubation vehicle.
• What must hold for the $850M Merge mark: that hemodynamic-signal latency can be engineered into a usable real-time interface AND first-in-human ultrasound-BCI efficacy lands AND ultrasound wins indications electrodes can't. Three independent "must-holds" on an unproven modality.
• −20–30% growth scenario: for a pre-revenue thesis, "growth disappoints" = the NHS readout underwhelms → Merge's next round flat/down 30–50% → the whole ultrasound narrative re-rates lower.
• Single permanent-impairment scenario: the hemodynamic-latency ceiling proves fundamental for closed-loop control. Plausibility: moderate-to-high for high-bandwidth use-cases; low for slow mood-state monitoring. That bifurcation is the investment question.

Lens 14 · Management Questions (ordered by information value)

1. What is the precise legal and IP relationship between Forest Neurotech (nonprofit) and Merge Labs (for-profit)? Was Forest IP licensed/assigned to Merge, on what terms, and how was that valued for the benefit of Forest's donors? (Highest-value: it determines whether there's anything left in Forest and whether the governance is clean.)
2. What is the actual closed-loop latency of Forest 1 in humans for a decoded command, and what is the engineering path below the ~2.4s hemodynamic floor?
3. For which use-cases is fUSI's hemodynamic signal sufficient (mood/state monitoring, neuromodulation) versus insufficient (real-time motor/communication control)? Where do you concede electrodes win?
4. What are the predefined success criteria for the NHS Forest 1 trial's mood/motivation endpoints, and what reads as a clear failure?
5. How stable is fUSI imaging through an implanted cranial window over months/years (scarring, window opacification, signal drift)?
6. Given the "no FDA in our lifetime" stance, what is the actual commercialization pathway for the validated technology — and is that pathway Merge, by design?
7. How single-source is the Butterfly dependency, and what happens to Forest 1 if Butterfly's going-concern or roadmap changes?
8. What is the safety profile of chronic focused-ultrasound neuromodulation — thermal/mechanical effects on tissue over repeated sessions?
9. What is Forest's remaining funded runway and its fixed-term FRO completion milestone — what does "done" look like for the nonprofit?
10. How do you manage the conflict of interest where the same founders lead the donor-funded nonprofit and the OpenAI-funded for-profit pursuing the same modality?
11. What spatial resolution / depth is actually achieved in humans (vs. the 100µm / 10x-fMRI primate figures)?
12. What is your response to electrode incumbents (Synchron, Neuralink) reaching pivotal/FDA milestones first in the high-value indications?
13. What molecular/biological approaches (the "molecular" half of Merge's pitch) complement ultrasound, and does that IP sit in Forest or Merge?
14. What would it take for Forest's technology to be licensed to multiple independent commercializers, not just Merge — i.e., is the nonprofit truly modality-agnostic about who benefits?
15. What is the single result in the next 18 months that would most change your own conviction in the ultrasound modality?

---