Venture Capital & Startup Funding Roundup, September 24, 2026: ARK Invest, Evolution Equity, Mirae Asset Capital, Socratic Partners & More

Venture capital today is clustering around companies that control a hard bottleneck rather than simply adding another application layer. The largest checks in the latest 12-hour window went to an enterprise-security browser, a brain-computer interface company, an alternative pharmacy-benefits platform, AI-driven drug discovery, and programmable financial infrastructure. That is a telling mix: investors are paying for control over browsers, neural interfaces, healthcare economics, proprietary biological data, and transaction rails — places where distribution, regulation, hardware, or hard-won data can create barriers that a new model release cannot erase.

The concentration is striking. The five largest dollar-denominated rounds in this selection — Island, Precision Neuroscience, Rightway, Basecamp Research and BigHat Biosciences — account for $1.02 billion. Across the ten companies below, disclosed financing totals $1.1175 billion plus C$17 million, and roughly 91% of the U.S.-dollar capital sits in those five largest deals. The message is not that early-stage venture has disappeared: Prime Minute raised a $15 million pre-seed and Axya a C$17 million Series A. It is that investors appear willing to write unusually large checks when a startup can plausibly own infrastructure, differentiated science or a defensible enterprise control point.

The Macro Environment: Capital Is Moving Toward Control Points

The wider venture market is no longer capital-starved as it was after the 2021 cycle. Crunchbase reported $510 billion of global startup investment in the first half of 2026, already above the $440 billion invested during all of 2025 and a record for any half-year in its dataset. AI remains a major driver, but today’s transactions show the money increasingly radiating outward from foundation models into the systems that AI makes more valuable: cybersecurity, optical networking, biotechnology, financial settlement, industrial procurement, and physical infrastructure.

That shift helps explain the size of several rounds. Island is not training a general-purpose model; it is building an enterprise browser as the security and policy layer through which employees and AI agents access corporate systems. Precision Neuroscience is financing a regulated, physical interface to the human brain. PicoJool is attacking optical-connectivity constraints inside AI infrastructure. BigHat is combining machine learning with experimental biology. These are capital-intensive propositions, but they also possess forms of defensibility that pure software companies often struggle to sustain.

Investor psychology also looks increasingly bifurcated. At the upper end, investors are prepared to finance private companies at scales once associated with pre-IPO rounds: Island’s $400 million Series F values it at $6.4 billion, while Rightway raised $155 million in Series E. At the other end, Eclipse and 8VC committed $15 million before Prime Minute had emerged from stealth. The middle of the market still exists, but the strongest enthusiasm is being expressed either very early — when technical and founder risk can be priced cheaply — or much later, when customer adoption and category leadership are easier to measure.

Another distinction is worth watching: AI is increasingly being financed as an input to a system rather than as the entire product. Rightway uses AI inside pharmacy-benefit operations; Axya applies it to manufacturing procurement; Trebellar is bringing vertical software and AI into corporate real estate; Prime Minute combines software with sensors, aviation and emergency-response operations. The model matters, but investors are also underwriting workflow ownership, proprietary datasets, physical systems, regulatory positioning and direct customer relationships.

Top Funding Rounds

Island raises $400M to turn the enterprise browser into an AI-era security control point

Island raised $400 million Series F funding led by Evolution Equity Partners. The round values the Dallas-based enterprise-browser company at $6.4 billion. The financing follows a $250 million Series E at a $4.8 billion valuation only months earlier and pushes the company’s cumulative financing to roughly $1 billion. Island plans to invest further in product development and international expansion across Europe, Asia, and the Middle East.

The investment is important because browsers are becoming more, not less, strategically valuable as companies deploy AI agents. Traditional enterprise security assumed a human employee moving between approved applications. Agents can operate across many services, execute actions at machine speed, and create another class of identity and data-governance problems. Island’s thesis is that the browser can become the enforcement point between users, agents, SaaS applications, and corporate information. A $6.4 billion valuation says investors believe that control point could become a large security category rather than a browser feature.

Funding Details
Startup: Island
Investors: Evolution Equity Partners led the round
Amount Raised: $400 million
Total Raised: Approximately $1 billion
Funding Stage: Series F
Funding Date: September 24, 2026
Headquarters: Dallas, Texas
Sector: Cybersecurity/enterprise browser / AI governance
Valuation: $6.4 billion

Precision Neuroscience raises $250M to push minimally invasive brain-computer interfaces toward commercialization

Precision Neuroscience closed an oversubscribed $250 million Series D co-led by Pershing Square Inc., the Ackman Oxman Institute and a life-sciences investment fund, bringing total capital raised to $430 million. Participants included Duquesne Family Office, B Capital, ARK Invest, Invus, Mubadala Capital, Mirae Asset Capital, Korea Investment Partners, Hitachi Ventures and JSL Health Capital. The company will use the financing to expand clinical work, move its Layer 7 cortical interface through further FDA review, and build commercial infrastructure.

Precision’s approach is a different engineering bet from brain implants designed to penetrate deep into neural tissue. Its Layer 7 interface is designed to sit on the brain’s surface; the company says it packs 1,024 electrodes into an area roughly the size of a postage stamp and is thinner than a human hair. Precision has already reported more than 100 patient procedures across 18 medical institutions, has FDA clearance for part of the platform, and has a commercial relationship with Medtronic. The $250 million round therefore looks less like speculative neuroscience financing and more like funding for the costly transition from clinical technology to a regulated medical platform.

Funding Details
Startup: Precision Neuroscience
Investors: Pershing Square Inc., Ackman Oxman Institute and a life-sciences investment fund co-led; Duquesne Family Office, B Capital, ARK Invest, Invus, Mubadala Capital, Mirae Asset Capital, Korea Investment Partners, Hitachi Ventures, JSL Health Capital and others participated
Amount Raised: $250 million
Total Raised: $430 million
Funding Stage: Series D
Funding Date: September 24, 2026
Headquarters: New York, New York
Sector: Neurotechnology / brain-computer interfaces / medical devices
Valuation: Not disclosed

Rightway raises $155M funding to challenge the economics of traditional pharmacy benefits

Rightway raised $155 million in Series E financing led by Francisco Partners, with existing investors Thrive Capital and Khosla Ventures participating. The New York company operates a pharmacy-benefit-management and healthcare-navigation platform and says it now serves 45 Fortune 500 companies. The financing will go partly toward AI and the technology underpinning its pharmacy-benefits operation.

The strategic issue isn’t whether AI can answer benefits questions. It is whether technology can alter the incentives and operating costs inside one of America’s most expensive healthcare middle layers. Rightway markets a model built around rebate pass-through, spending guarantees, pharmacists, and navigation tools rather than the economics associated with traditional PBMs. Investors are effectively underwriting the possibility that large employers will switch from incumbents when a challenger can demonstrate both lower drug spending and a credible member experience. That is a much larger prize than selling HR software on a per-seat basis.

Funding Details
Startup: Rightway
Investors: Francisco Partners led; Thrive Capital and Khosla Ventures participated
Amount Raised: $155 million
Total Raised: Not disclosed in the announcement
Funding Stage: Series E
Funding Date: September 24, 2026
Headquarters: New York, New York
Sector: Healthtech / pharmacy benefits/care navigation
Valuation: Not disclosed

Basecamp Research raises $140M to train AI systems for drug discovery

Basecamp Research announced a $140 million Series C financing. The company describes itself as a frontier-AI business developing AI-designed medicines, with the capital intended to train a new generation of biological models. The accessible correction notice confirmed the round size, stage, and focus, but did not disclose the full investor syndicate in the material available during this verification pass.

The investment belongs to one of the more interesting branches of AI funding: companies trying to build biological intelligence from differentiated training data rather than relying solely on publicly available protein and genomic datasets. For this class of business, the defensibility question is unusually concrete. Better algorithms help, but the enduring asset may be proprietary biological data connected to experimental and drug-development workflows. Investors backing large rounds in this category are betting that model quality in biology will increasingly depend on the uniqueness and scientific relevance of the underlying dataset, not simply compute scale.

Funding Details
Startup: Basecamp Research
Investors: Full syndicate not available in the accessible correction notice reviewed
Amount Raised: $140 million
Total Raised: Not disclosed in the source reviewed
Funding Stage: Series C
Funding Date: September 24, 2026
Headquarters: London, United Kingdom
Sector: AI drug discovery / computational biology
Valuation: Not disclosed

BigHat Biosciences raises $75M to pair AI protein design with experimental biology

BigHat Biosciences announced a $75 million Series C co-led by DFJ Growth and Premji Invest. The San Mateo biotech says the financing will support both its investigational drug pipeline and its protein-design data platform; current reporting puts cumulative funding at roughly $223 million. BigHat is already in the clinic with BHB810, an AI-designed antibody-drug conjugate being evaluated in gastrointestinal cancers.

What makes the model financially interesting is the closed loop between software and laboratory evidence. Protein-design models can propose candidates rapidly, but biological experiments determine whether those candidates bind, behave safely, and possess properties suitable for a real medicine. Owning both sides of that iteration can generate proprietary experimental data that feeds the next design cycle. The harder question is whether that advantage survives clinical translation. At Series C, BigHat’s investment case is increasingly tied to human readouts rather than model benchmarks — which is exactly where an AI-biotech company should ultimately be judged.

Funding Details
Startup: BigHat Biosciences
Investors: DFJ Growth and Premji Invest co-led the round
Amount Raised: $75 million
Total Raised: Approximately $223 million
Funding Stage: Series C
Funding Date: September 24, 2026
Headquarters: San Mateo, California
Sector: Biotechnology / AI protein design/therapeutics
Valuation: Not disclosed

HIFI raises $37M to build programmable infrastructure for tokenized money

New York-based HIFI raised a $37 million Series A led by Left Lane Capital. The company provides infrastructure that connects bank rails, stablecoins, and tokenized assets through APIs. HIFI says its systems now handle more than $7 billion of annualized volume and have been used to onboard more than 10,000 businesses and 200,000 individuals. The round will fund additional regulatory licenses, expansion in New York and selected international markets, and moves beyond payments into cards and capital-markets infrastructure.

The financing illustrates how crypto investment has changed. The venture thesis here is less about speculative token prices and more about settlement architecture. HIFI has participated in tokenized-repo activity involving DTCC infrastructure and supported a live transaction between DRW and Marex on the Canton Network. If tokenized cash, securities and collateral gain institutional adoption, companies that abstract away custody, compliance, banking connections and cross-network settlement may occupy a valuable position between traditional finance and blockchain-based rails.

Funding Details
Startup: HIFI
Investors: Left Lane Capital led
Amount Raised: $37 million
Total Raised: $37 million reported by Axios
Funding Stage: Series A
Funding Date: September 24, 2026, 8:00 a.m. ET
Headquarters: New York, New York
Sector: Fintech/stablecoin infrastructure / tokenized capital markets
Valuation: Not disclosed

PicoJool raises $27.5M to attack the optical bottleneck inside AI clusters

Palo Alto-based PicoJool raised a $27.5 million Series A led by Socratic Partners with participation from Hudson River Trading. The round follows a $12 million seed led by Playground Global, taking disclosed funding to $39.5 million. PicoJool develops high-speed microVCSEL optical technology, including 200G and 50G products, and plans to expand engineering, commercialization, and operations in the United States and Taiwan.

This is a relatively modest check compared with today’s late-stage rounds, but it may be strategically more important. AI-compute economics are becoming a system problem. A GPU is only useful if data can move between chips, racks and clusters quickly enough and within an acceptable power budget. That makes optical interconnects part of the AI infrastructure race alongside memory, advanced packaging, networking and power delivery. PicoJool is therefore competing for capital in a market where a successful component can ride the spending of multiple accelerator vendors and hyperscalers rather than betting on one model company.

Funding Details
Startup: PicoJool
Investors: Socratic Partners led; Hudson River Trading participated; Playground Global led the prior seed round
Amount Raised: $27.5 million
Total Raised: $39.5 million
Funding Stage: Series A
Funding Date: September 24, 2026, 8:00 a.m. EDT
Headquarters: Palo Alto, California
Sector: Semiconductors / optical interconnects / AI infrastructure
Valuation: Not disclosed

Trebellar raises $18M in funding to bring vertical AI into corporate real estate

Trebellar raised an $18 million Series A led by Blossom Capital, according to an Axios exclusive published today. The company is building a corporate real estate management platform for enterprises that still coordinate property portfolios through a mix of spreadsheets, static software and outside consultants.

The investment fits a broader enterprise-software pattern: venture firms are increasingly interested in AI products that own an entire professional workflow rather than a thin assistant sitting beside an incumbent system. Real estate is especially attractive because large companies generate large volumes of lease, occupancy, maintenance, energy, and facilities data, while decision-making remains fragmented across finance, workplace, and property teams. Trebellar’s challenge will be proving it can become the operating system for those processes rather than an intelligence layer that established property-software vendors can reproduce.

Funding Details
Startup: Trebellar
Investors: Blossom Capital led
Amount Raised: $18 million
Total Raised: Not disclosed in the source reviewed
Funding Stage: Series A
Funding Date: September 24, 2026
Headquarters: United States; city not specified in the source reviewed
Sector: Proptech / enterprise software / vertical AI
Valuation: Not disclosed

Prime Minute raises $15M to build technology for wildfire and disaster response

Prime Minute emerged from stealth with a $15 million pre-seed led by Eclipse with participation from 8VC. The Palo Alto company is developing an integrated disaster-response system that combines real-time information, AI-assisted mission analysis, hardware and aerial wildfire-suppression capabilities. The capital will go toward product development, domestic manufacturing, hiring, testing and validation. Prime Minute says it is working with Colorado on prototype field evaluations later this year.

A $15 million pre-seed is large because the problem demands far more than an enterprise-software launch. Prime Minute has to integrate sensing, mission software, aerial systems, manufacturing and real-world public-safety operations. That puts the company near the intersection of defense technology, climate adaptation and autonomous systems — a category increasingly attractive to investors that have learned from defense startups that public-sector problems can support venture-scale companies when software and hardware are designed together. The commercial risk is equally clear: reliability, procurement, and field performance matter far more than demo quality.

Funding Details
Startup: Prime Minute
Investors: Eclipse led; 8VC participated
Amount Raised: $15 million
Total Raised: $15 million
Funding Stage: Pre-seed
Funding Date: September 24, 2026, 9:00 a.m. ET
Headquarters: Palo Alto, California
Sector: Disaster-response technology/climate adaptation / autonomous systems
Valuation: Not disclosed

Axya raises C$17M to automate industrial procurement with AI

Montreal-based Axya raised C$17 million in Series A funding, led by McRock Capital, with Yamaha Motor Ventures participating alongside existing shareholders including BDC’s Industrial Innovation Venture Fund and Real Ventures. The company builds procurement software for manufacturers that automates parts of sourcing and purchasing, and says the financing will expand its AI capabilities, geographic reach, engineering team, and go-to-market operation.

Manufacturing procurement is a useful example of why investors continue funding vertical AI despite the flood of general-purpose models. Industrial buyers must interpret specifications, qualify suppliers, compare quotations, handle production constraints and preserve institutional knowledge about vendors and parts. A model alone does not own those relationships or transaction histories. A platform embedded in the procurement process can. Yamaha Motor Ventures’ participation is particularly relevant because strategic industrial investors can provide more than capital: they can help validate whether software survives the demands of actual manufacturing environments.

Funding Details
Startup: Axya
Investors: McRock Capital led; Yamaha Motor Ventures, BDC Industrial Innovation Venture Fund, Real Ventures and existing shareholders participated
Amount Raised: C$17 million
Total Raised: Not disclosed in the announcement
Funding Stage: Series A
Funding Date: September 24, 2026, 9:30 a.m. ET
Headquarters: Montreal, Quebec, Canada
Sector: Industrial software/manufacturing procurement / AI
Valuation: Not disclosed

What Today’s Funding Activity Reveals

Capital is moving below the application layer. PicoJool is a particularly clean example: the company is not competing to build a chatbot or model. It is working on the optical infrastructure needed to connect the machines running those models. HIFI is building settlement infrastructure beneath financial applications. Precision is creating the physical neural interface through which future software could interact with the brain. BigHat is building an experimental-data loop beneath AI-assisted drug design. These businesses may have longer development cycles, but their value is tied to difficult engineering and infrastructure rather than a transient model advantage.

AI security is shifting from protecting employees to governing employees plus agents. Island’s financing is the strongest signal in the group today. A $400 million round at $6.4 billion implies investors see enterprise browsing as a much broader control surface as autonomous software takes on more work. Security products will increasingly need to answer not just whether a user is authorized, but what an agent acting for that user may read, transmit, execute or modify.

Healthcare capital is flowing toward businesses that combine technology with a hard external proof point. Precision must show regulatory progress and clinical utility. BigHat’s compounds ultimately have to work in patients. Rightway must demonstrate measurable employer savings. Basecamp’s biological models have to produce drug candidates that survive laboratory and clinical validation. The common feature is that a persuasive software demo is insufficient. That raises execution risk, but it can also create deeper defensibility if the company succeeds.

The United States remains dominant in this particular 12-hour slice, but the capital base is international. Eight of the ten selected companies are U.S.-headquartered or primarily U.S.-based, while Basecamp is U.K.-based and Axya is Canadian. The investor mix, however, reaches further: Mubadala Capital and Asian investors appear in Precision’s syndicate; PicoJool is expanding in Taiwan; BigHat attracted Premji Invest; and Axya brought in the venture arm of Japan’s Yamaha Motor. Today’s flow therefore looks less like national venture silos and more like global pools of capital underwriting companies positioned around U.S.-led technology spending.

Round size is becoming a signal of category conviction. Island, Precision and Rightway together raised $805 million. By contrast, technically interesting Series A and pre-seed businesses received tens of millions. That is rational portfolio behavior in a market where investors can make large commitments to companies with stronger evidence while preserving smaller option-like bets on emerging infrastructure categories. It also means headline venture totals can look exceptionally strong even when capital remains selective.

Venture Funding Table

Strategic Takeaways for Founders and Investors

For founders, today’s clearest message is that simply attaching AI to an existing workflow is unlikely to be enough. The companies attracting serious money tend to own something deeper: Island controls the browser where work occurs; HIFI owns transaction plumbing; PicoJool is building hardware; BigHat generates experimental biological data; Precision owns a regulated neural interface; Rightway controls both software and an economic model for drug benefits. Defensibility comes from what accumulates around the AI—data, compliance, hardware, distribution, customer integration, and domain-specific operations.

That creates a practical fundraising lesson. AI should increasingly be presented as leverage, not as the moat itself. Investors can assume model capability will improve and costs will fall. A founder therefore needs to answer the harder question: what becomes more valuable as models commoditize? In Axya’s case, it could be supplier and purchasing workflows. For Trebellar, it could be the system of record around enterprise property operations. For Prime Minute, it is the marriage of software with field hardware, manufacturing and emergency-response execution.

Capital efficiency also needs to be judged by sector rather than by a universal software metric. A $15 million pre-seed may sound heavy for a SaaS business, but Prime Minute has physical systems to build and test. Precision’s $250 million Series D is enormous by enterprise-software standards, yet medical-device trials, manufacturing, and FDA processes demand substantial capital. Conversely, a pure software startup raising at those levels without equivalent technical, regulatory or distribution barriers would face a much harder burden of proof.

For investors, the more interesting question is where AI creates new scarcity. Compute scarcity has already pushed money into chips, networking and power. Agent proliferation creates scarcity in trusted identity, permissioning and security, helping explain Island. Tokenized assets create demand for reliable settlement and compliance infrastructure, helping explain HIFI. More powerful biological models increase the value of proprietary experimental data, which supports the investment logic around Basecamp and BigHat.

Pricing power will probably accrue to companies sitting at those scarce control points. A generic AI assistant faces substitution from foundation-model vendors, open-source software, and competitors using the same APIs. A platform responsible for moving regulated money, controlling access to corporate systems, qualifying industrial suppliers, producing validated biological data or interfacing safely with the brain is harder to switch. That does not guarantee good venture returns — valuations still matter — but it provides a stronger basis for durable gross margins and customer retention than model novelty alone.

Finally, founders should pay attention to who is leading rounds. Today’s list includes specialist or operationally oriented investors such as Evolution Equity Partners in cybersecurity, Francisco Partners in technology-enabled healthcare, Eclipse in physical industries, McRock in industrial technology, and a growth investor pairing in BigHat. Sector knowledge becomes more valuable as startup execution moves from shipping software toward regulatory approvals, manufacturing, healthcare economics, and industrial deployment.

Conclusion

Today’s funding flow is not primarily a story about another wave of AI applications. It is a story about the infrastructure, control points, and proprietary assets forming around AI.

Island is being financed to police the browser through which humans and agents work. PicoJool is being financed because AI systems need faster optical connections. BigHat and Basecamp are applying machine intelligence where proprietary biological data can compound. Precision Neuroscience is connecting computation to the human nervous system. HIFI is rebuilding parts of financial settlement for programmable money. Prime Minute is taking software into wildfire response. Axya and Trebellar are embedding it into industrial procurement and corporate real estate. Rightway is applying technology to one of healthcare’s most expensive economic choke points.

The common denominator is not AI by itself. It is AI attached to something difficult to reproduce: regulated infrastructure, physical hardware, experimental data, transaction networks, clinical relationships, industrial workflows, or deeply embedded security controls.

That distinction may define the next phase of venture investing. As general intelligence becomes cheaper and more widely available, the scarce assets surrounding it become more valuable. Today’s strongest funding rounds suggest that investors are already positioning accordingly.