In October, private companies globally raised 1,350+ early-stage rounds globally (noting this total will rise as more deals are published retroactively).

Download the full report to access comprehensive CB Insights data on early-stage activity, including top investors & deals, valuation data, and our predictive signals. Below, we highlight notable trends to watch.

Review the full deal dataset

Emerging tech trends to watch

Click the links to see underlying deal activity. Companies mentioned based on CB Insights Mosaic startup potential scores and relevance.

Stablecoin infrastructure is blockchain’s breakout category

Blockchain startups raised over 80 deals in October (out of 1,394 total early-stage deals), in line with September’s activity. The deals accounted for $1.1B in funding, representing 13% of total early-stage dollars invested in the month and a major jump from prior months, thanks to Tempo’s $500M round. 

Overall, institutional investor activity (20 blockchain deals had CVC backers) and a median deal size of $6M (vs. $2.5M across the venture landscape) indicates investor conviction in a maturing blockchain landscape. See the companies already deploying their solutions in the market with CB Insights Commercial Maturity scores here.

Stablecoin infrastructure emerged as a leading theme, with 15+ companies raising capital in October. 

The build-out is happening across multiple layers. At the protocol level, Tempo (744 Mosaic), a Layer-1 blockchain developed by Stripe and Paradigm for stablecoin transactions and payments, raised the largest round of the month — a $500M Series A at a $5B valuation. 

Activity extends to the middleware layer, with Tesser (618 Mosaic, $4.5M seed) and Cybrid (705 Mosaic, $10M Series A) building API infrastructure that lets traditional banks integrate stablecoin capabilities.

Geographic diversification is notable as well. Lumx (596 Mosaic, $3.4M seed) focuses on stablecoin payments infrastructure in Latin America, while Standard Economics (671 Mosaic, $9M) is targeting developing markets with cross-border payment solutions. 

The activity signals the potential of stablecoin rails to become core financial infrastructure. 

Review the stablecoin market map (May 2025) here.

AI agents make inroads across semiconductor design, healthcare, payments

Companies targeting AI agent applications & infrastructure raised over 70 deals in October. Smart Money investors — the top 25 VCs identified by CB Insights — backed 16 AI agent startups (22%) in the month.

Key emerging trends to note based on deal activity include:

• Browser automation & web agent infrastructure (5 deals): Most business-critical workflows still live in web UIs that require manual navigation rather than APIs. Companies like Kernel (646 Mosaic, $22M Series A) provide browsers-as-a-service infrastructure that enables agents to interact with websites.
• Semiconductor & hardware design (5 deals): Hardware engineering, from semiconductor chips to mechanical CAD, involves complex, time-intensive design and verification tasks. ChipAgents (710 Mosaic, $21M Series A) automates chip design workflows, while Adam (576 Mosaic, $4.1M) provides AI-powered CAD tools for mechanical engineering and 3D design.
• Voice agents for healthcare (4 deals): Healthcare providers face labor shortages and 24/7 patient communication demands. Attuned Intelligence (669 Mosaic, $13M seed), Remedy (474 Mosaic, $0.5M YC note), and OutcomesAI ($10M seed) deploy voice agents that handle appointment scheduling, prescription refills, and patient triage autonomously.
• Payments & transaction infrastructure (4 deals): To enable agents to initiate transactions, companies like Kite AI (702 Mosaic, Series A) are building specialized payment infrastructure. Others in this category such as Pagentic (562 Mosaic, $2M pre-seed) are focused on building payments and billing infrastructure for monetizing AI agents’ activities.

Compare 17 players in the AI agent payments infrastructure landscape on CB Insights.

Real-world training data providers scale specialized AI capabilities

Foundation models are bottlenecked by training data quality and diversity. October early-stage activity (6 deals) highlights specialized data providers emerging as important infrastructure for AI development, particularly for reasoning, coding, and physical-world applications. 

For example: 

• DataCurve (822 Mosaic, $15M Series A) offers a “bounty” platform where software engineers complete coding tasks to generate high-quality training data for models
• General Intuition (664 Mosaic, $134M seed) is training spatial-temporal reasoning models on interactive video data from gaming environments
• Lucent (507 Mosaic, $1.3M pre-seed) creates behavioral datasets from real-world product web interactions to train browser agents
• Hillclimb (479 Mosaic, $0.5M YC note) connects mathematicians with AI labs to create advanced math training data to improve models’ reasoning abilities

Wildfire detection & prevention technology attracts investment

Wildfire seasons are intensifying globally, driving demand for new detection and response technologies (5 deals). Companies like SenseNet (Mosaic 777, $10M Series A) and Frontline Wildfire Defense (Mosaic, $48M Series A) are deploying detection networks (sensors, cameras, and satellite data), automated suppression systems (sprinklers, foam, drones), and predictive risk platforms for utilities, governments, and property owners in fire-prone regions.

Review the wildfire tech market map (February 2025) here. 

Methodology

This report includes equity early-stage financings (convertible note, angel, pre-seed, seed, Series A) to private companies in October 2025. We excluded companies that are later-stage that raised an angel round or convertible note in the month. Categorization based primarily on company descriptions.

FURTHER READING

• State of Venture Q3’25
• State of AI Q3’25
• Early-Stage Trends Report: Smart Money is all in on AI agents, the rise of autonomous labs, and more in September
• The AI agent market map

The post Early-Stage Trends Report: Stablecoins’ breakout moment, real-world AI training data, and more in October appeared first on CB Insights Research.

These companies are seeing success as enterprises realize ROI through reduced need for human support staff, and lowering costs and increasing revenue through improved operational efficiency. These results explain why the market rates highly on our Mosaic Score, averaging 714, placing it in the top 5% of all markets in the enterprise tech industry.

So who’s leading today, who’s gaining ground fastest?

Using CB Insights’ revenue data, we identified each company’s customer service AI agents & copilots revenue to measure the current market size and estimate market shares for private players in the space.

If you are active in the customer service AI agents & copilots market and want to submit your company’s revenue data, please reach out to researchanalyst@cbinsights.com. 

See all customer service AI agents & copilots market ranked by revenue

Key takeaways

• New entrants are gaining ground rapidly using generative AI technology, to challenge leaders. Companies like Sierra, Crescendo, Decagon, Chatbase, founded in the last 2 years have already broken into the top 10 by revenue generation, competing directly with legacy players. This highlights low barriers to achieving scale in customer service AI where superior AI models and user experience can quickly displace incumbents who are reacting by acquiring some of these startups to stay competitive. For example, ServiceNow acquired Moveworks for $2.85B while Cognigy was bought out by NiCE in a $955M deal.
• The market is rapidly transitioning to autonomous agents that resolve customer issues end-to-end without human escalation. This is disrupting traditional pricing models with vendors like Intercom, Sierra and Decagon now charging per successful resolution rather than per seat or usage. Enterprise adopters must ensure any tool they adopt includes performance validation capabilities that link AI spending directly to resolution success and cost reduction metrics.
• Value of enterprise adoption will come from integration depth. AI agents require seamless integration to company-specific context such as historical tickets, internal knowledge bases, CRM data, and proprietary documentation, to deliver personalized, informed resolutions while maintaining consistent brand voice. Enterprise value emerges when agents turn customer inquiries into closed-loop resolutions, eliminating inefficient human handoffs and workflow bottlenecks. According to CB Insights buyer reports, successful vendors offer agentic solutions with ease of deployment and integration.

• Agentic voice AI expands the value proposition. Advances in voice AI, from natural dialogue and real-time tone analysis to sentiment-based adaptation, position voice as the emerging primary interface between humans and technology. By capturing emotional cues and conversational nuance that text misses, agentic voice systems can detect upsell opportunities, flag churn risks, and surface product insights during live interactions, transforming voice support from a cost center into a revenue driver. As these capabilities mature, enterprises should shift KPIs from operational efficiency (e.g., time or cost savings) toward strategic outcomes such as customer retention, conversion lift, and topline revenue growth directly attributable to AI-enabled engagement.

Market Overview

The customer service AI agents & copilots market comprises tools that automate client support inquiries and processes across multiple communication channels. These solutions converse with customers, answer questions, support inquiries, and resolve issues. Many can handle complex inquiries and complete processes end-to-end with minimal human oversight. They may also provide real-time assistance to human agents.

The market includes a mix of commercially mature players (~40%), challengers like Decagon and Crescendo, and established customer service incumbents. They have raised a combined $1B in equity funding so far this year, already matching last year’s amount.

The post The customer service AI agents leading the market in 2025 appeared first on CB Insights Research.

However, unlike September’s concentrated focus on raw compute and model training, October’s AI infrastructure deals spread beyond models to encompass the full stack, from semiconductors to open-source databases.

This funding distribution presents a clear pattern: investors are backing companies that help businesses use AI, not just build it. This includes agent infrastructure, workflow automation, and verticalized AI tools for legal, healthcare, and industrial sectors.

Using CB Insights’ predictive signals, our monthly Book of Scouting Reports offers an in-depth analysis of every private tech company that has raised a funding round of $100M or more. It spotlights where private capital is concentrating, startups gaining momentum, and which companies are becoming tomorrow’s AI disruptors.

Download the book to see all 48 scouting reports.

The post Beyond foundation models: October mega-rounds span the full AI infrastructure stack appeared first on CB Insights Research.

Early signals in hiring and partnerships show the compute mix starting to split. Lead times, cost, and concentration risk are pushing teams to add a second path. Availability of compute remains tight, and this is driving capacity to widen beyond a single vendor.

Anthropic is an early test case. The company is helping to develop Amazon’s Trainium AI chips, including plans to use these chips to train its next Claude model, and is expanding its use of Google’s TPU chips. These moves add capacity beyond Nvidia and lessens its reliance on Nvidia hardware.

Using CB Insights customer sentiment interviews, hiring insights, investments, and partnership activity, we analyze:

• Why Nvidia dominated and what’s different now
• Anthropic as an early example of diversification
• Beyond Anthropic, signals of a multi‑chip market

Why Nvidia dominated and what’s different now

Nvidia’s historic dominance in the AI chip space stems from its software and hardware ecosystem, and while that foundation remains strong, alternatives have improved enough to enable hedging.

CUDA — Nvidia’s software platform and programming model for running AI on its GPUs — and its supporting toolchain compress time to adopt and reduce delivery risk. For most teams, swapping to another supplier means rewriting code, retraining people, and revalidating models, a tax few companies can absorb, especially when developing new AI solutions.

CB Insights customer sentiment interviews echo this lock-in. For example, one founder at a software engineering company mentioned “since all of our infrastructure is built on CUDA … the main challenge in switching … is the infrastructure and software ecosystems. The learning cost … is unknown, and that uncertainty is a barrier.”

Check out more Nvidia customer sentiment interviews

And because CUDA software requires specialized knowledge, this lock-in shows up in AI company hiring data. For example, Baseten is hiring a GPU Kernel Engineer to implement custom CUDA kernels across current‑gen Nvidia GPUs. That is a direct signal of Nvidia‑specific optimization work that raises near‑term switching costs.

But parts of this story are changing. AMD’s software has materially improved since late 2024. Issues that once blocked deployments are less frequent, and fixes land faster. Parity is not universal by workload, but the friction gap is narrowing enough to allow some teams to start diversifying.

Anthropic as an early example of diversification

Against this backdrop, Anthropic is the early test of real diversification.

Anthropic is the notable exception to Nvidia’s investments across leading AI companies. The two have recently clashed over export controls, signaling a strained relationship between Nvidia and Anthropic that continues to this day.

See Nvidia’s latest AI investments here

The company’s relationship with Amazon further signals a path away from a sole reliance on Nvidia. This relationship began in September 2023, when Amazon announced a $4B investment in Anthropic and, in turn, the company named AWS its primary cloud provider. This meant distribution of Anthropic’s Claude models through AWS’s Bedrock service, as well as access to the Trainium AI chips that Amazon has been developing since 2020.

Since then, the two appear to be working in tandem to diversify away from Nvidia chips. AWS, for instance, recently said more than half of its Bedrock service now runs on its custom Trainium AI chips. In addition, Anthropic has recently committed to training the next generation Claude model on Trainium chips through Amazon’s Project Rainier data center deployment.

And Anthropic isn’t just relying on Amazon to help it move away from Nvidia chips.

In October 2025, Anthropic announced it will expand its use of Google TPUs — Google’s specialized AI chips — to up to one million units. Until then Google kept most TPU capacity for internal use; opening it at Anthropic scale signals both Anthropic’s push to add non‑Nvidia supply and Google’s bet that more buyers want alternatives. That shift on both sides sets up the broader signals of a multi‑chip market.

Beyond Anthropic, signals of a multi‑chip market

As software improves and compute remains hard to source, adopting alternatives is easier than a year ago. Signals include earnings calls, hiring, and partnerships.

In recent earnings calls, Amazon has highlighted persistent AI capacity constraints resulting in a strategic push to develop its Trainium chips. The company notes that AI demand is outpacing supply, opening up opportunities to develop its own capacity through Trainium chip supplies. But these supplies are still limited, forcing buyers to hedge delivery risk by lining up other non‑Nvidia capacity. For Anthropic, that likely meant adding Google TPUs alongside Trainium.

Monitor Amazon’s Trainium chip development through Earnings Insights

Hiring signals also show the trend gaining traction with other AI startups:

• Nscale, a data center hyperscaler, is hiring core infra roles that require “experience with Nvidia and AMD.” Nscale is positioning itself as a primary EU compute supplier, targeting data‑local workloads and a diversified supply chain that may shift some regional AI compute away from Nvidia.
• TensorWave, on the other hand, uses explicit AMD‑powered positioning (“building the Infrastructure for AMD‑Powered AI Computing”) and mentions that the company has deployed North America’s largest AMD GPU cluster. The company is projecting $100M in ARR by year’s end, signaling the market’s appetite for non-Nvidia compute.
• Groq, an infra provider developing its own learning processing unit (LPU) chip to compete with conventional Nvidia/AMD GPUs, is still recruiting to “support evolving Nvidia and AMD GPU architectures,” indicating cross‑vendor software support even as it promotes its own LPUs.

Beyond hiring and capacity, anchor deals are pushing diversification. OpenAI is lining up multi‑GW compute across Nvidia, AMD, and Broadcom, with Broadcom focused on custom, jointly designed chips. Oracle — a key OpenAI compute partner — plans roughly 50,000 next‑gen AMD accelerators starting in Q3’26 with expansion through at least 2027, reinforcing the move away from a single‑supplier model.

For the market overall, as reliance on Nvidia eases at the margin, compute supply spreads across Nvidia, AMD, and cloud custom silicon. That expands access, improves tokens‑per‑second and latency, and reduces lead‑time risk for startups. Where throughput and latency improve, usage rises and spend follows — especially for teams already maxed on capacity.

As compute comes from multiple vendors, expect a larger market for software that abstracts chip differences so teams can focus on product rather than implementation.

If you are a venture investor and want to submit data on your portfolio companies to allow us to better score you in the future, please reach out to researchanalyst@cbinsights.com.

RELATED RESOURCES FROM CB INSIGHTS:

• All 100 AI unicorns since ChatGPT launched
• State of Tech Exits H1’25
• The top angel investors in AI
• Are AI unicorns starting to move beyond hype?
• Book of Scouting Reports: Enterprise AI Agents

For information on reprint rights or other inquiries, please contact reprints@cbinsights.com.

The post Partnerships and hiring data show AI companies are expanding beyond Nvidia chips appeared first on CB Insights Research.

To create the Tech IPO Pipeline, we scored companies across CBI datasets including Mosaic scores, hiring insights, revenues, exit probabilities, business relationships, and more.

Check out key highlights across the Tech IPO Pipeline below.

CBI customers: See these companies ranked by Mosaic and revenue

Combining CB Insights’ proprietary data and AI, scouting reports provide insight into each company’s:

• Funding history
• Headcount
• Key opportunities and threats
• IPO prospects
• Mosaic score

For customers, get the full book of 100+ scouting reports using the download button on the lefthand side.

MORE CB INSIGHTS RESEARCH:

• State of Venture Q3’25 Report
• Future Tech Hotshots: 2025: 45 emerging tech startups poised to make an outsized impact
• State of AI Q3’25 Report

For information on reprint rights or other inquiries, please contact reprints@cbinsights.com.

The post Tech IPO Pipeline 2026: Book of Scouting Reports appeared first on CB Insights Research.

Combining CB Insights’ proprietary data and AI, scouting reports provide insight into each company’s:

• Funding history
• Headcount
• Key takeaways (including opportunities and threats)
• Product/tech focus
• Mosaic score

Want to see more research? Start your free trial.

If you’re already a customer, log in here.

The post Book of Scouting Reports: Top 100 European AI Aerospace & Defense Companies appeared first on CB Insights Research.

The conflict in Ukraine demonstrates the significant impact of these technologies — think low-cost drones destroying $5M tanks — driving European investment in domestic defense systems and supply chains.

Want to see more research? Start your free trial.

If you’re already a customer, log in here.

The post The top 100 European AI aerospace & defense startups appeared first on CB Insights Research.

While deals fell in Q3’25, billion-dollar rounds to AI infrastructure players continued to drive the funding surge. But the activity isn’t limited to the largest players: investors are cutting bigger checks across every stage, signaling both conviction in AI’s potential and the high costs of AI development.

Among emerging opportunities, AI agents are a key focus for VCs and enterprises alike, with agent markets leading deal and M&A activity in the quarter.

Below, we break down the top stories from this quarter’s report, including:

• AI deal activity softens, but massive rounds support continued funding boom
• Consolidation remains in full force in the AI market
• Tech market deals highlight AI agent applications, rise of GEO
• The talent premium: AI companies valued at up to ~$100M per employee

Download the full report to access comprehensive CB Insights data and charts on the evolving state of AI across geographies.

AI deal activity softens, but massive rounds support continued funding boom

Deals to private AI companies globally fell 22% quarter-over-quarter in Q3’25, but funding remained above $45B for the fourth consecutive quarter.

Taken together, these trends indicate how top-heavy the AI venture funding landscape has become. 

The average deal size in 2025 YTD is $49.3M — up 86% from 2024. In the last 4 quarters, mega-rounds ($100M+ deals) have accounted for 75%+ funding. The average since 2021 (up to Q3’24) is 53%. 

At the same time, check sizes are trending bigger at the median across every stage this year. For example, the median early-stage deal is $3.4M in 2025 YTD, up from $2.5M in 2024. 

Investors are funneling capital into fewer, larger bets on perceived AI winners, driven by the massive infrastructure costs and competitive dynamics of foundation model development.

In Q3’25, there were 6 $1B+ rounds alone. The top 3 deals went to LLM developers — Anthropic ($13B, Series F), OpenAI ($8.3B, PE), and Mistral AI ($1.5B, Series C) — reflecting the high cost of frontier model development. While OpenAI hit $12B in annualized revenue in July 2025, it’s projecting roughly $8B in cash burn this year per reports. 

Other infrastructure players like Nscale (AI data centers, $1.1B Series B) and Groq (AI inference processors, $750M, Series E) were also in the top 10. The raises are indicative of the growth and attention technologies enabling AI are receiving, with earnings call mentions of data centers hitting record levels in Q3’25 and AI training & inference chips on track for record equity deal & funding activity this year.

Consolidation remains in full force in the AI market

The AI market is a hotbed for M&A activity. 

Q3’25 marks the second highest quarter on record for AI startup M&A (172 deals), following Q2’25 (181 deals). The US continues to gain share, with startups based in the country accounting for 59% of total exits, the highest share since Q2’21. 

Three of the top 5 exits in the quarter were related to AI agents: 

• Workday acquired Sana Labs, a company focused on enterprise workflows, for $1.1B. 
• NiCE acquired customer support startup Cognigy for $955M. 
• Atlassian acquired The Browser Company, the maker of the Arc and Dia AI browsers, for $610M. 

The activity signals enterprise software incumbents are looking to buy their way into accelerating their AI roadmaps. Workday was the second most active acquirer in the quarter with 3 acquisitions (behind Salesforce, with 4 acquisitions). The HR & finance software company also picked up agent builder Flowise and AI-powered recruiting platform Paradox.

Meanwhile, Meta made its first publicly disclosed acquisitions since 2022, acquiring voice AI startups Play AI and WaveForms AI.

Other notable top exits include AI security companies Lakera (acquired by Check Point for $300M) and Prompt Security (acquired by SentinelOne for $250M-$300M). Generative AI is expanding attack surfaces, driving large cyber players to opt for M&A to more quickly integrate AI security features into existing offerings.

Both Lakera and Prompt Security were founded less than 5 years ago, far “younger” than the average time to exit of 9.7 years in the quarter, underscoring how rapidly AI security has become mission-critical.

Review the AI security startups that are ripe for acquisition next in this brief.

Tech market deals highlight AI agent applications, rise of GEO

Among the 1,500+ tech markets that CB Insights tracks, those in the chart below saw the greatest number of AI deals in Q3’25 (note: companies may appear in multiple markets).

Industrial humanoid robot developers and coding AI agents & copilots remained at the top, while LLM developers also climbed back up in the rankings from Q2’25.  

One notable rising market is generative engine optimization (GEO), which refers to tools that help brands optimize their visibility in AI search platforms like ChatGPT and Perplexity. This emerging category (the most nascent in the list based on CBI Commercial Maturity scores) addresses the shift toward shopping and discovery happening on top of LLM interfaces.

OpenAI’s September 2025 launch of in-platform shopping capabilities in ChatGPT underscores this trend, establishing AI platforms as new commerce channels requiring specialized optimization strategies.

Using CB Insights’ Mosaic score — which measures private company health and predicts likelihood of success — we analyzed more than 20 GEO companies, ranking them by 1-year Mosaic score growth to identify the fastest-rising vendors. 

See the GEO partners best positioned to help brands win in AI search here. 

The talent premium: AI companies valued at up to ~$100M per employee

AI companies with lean headcounts and breakthrough potential are attracting sky-high valuations.

Humanoid robotics developer Figure leads the pack in Q3’25 at $104.3M per employee on a $39B valuation, despite reporting no revenue last year (though projecting $9B by 2029). Cognition follows with $98.1M per employee, based on its $10.2B valuation. While the coding AI agent startup has $150M+ in ARR (following its acquisition of Windsurf), this indicates a lofty revenue multiple of ~68x. 

Others topping the quarter’s valuation-per-employee list span the AI model (Anthropic, Mistral AI, Decart, Harmonic), infrastructure (Baseten), and application layers (OpenEvidence,  Sierra, Irregular). 

Whether these valuations prove prescient or overextended will largely depend on whether these companies can deliver on ambitious revenue projections in the years ahead.

RELATED RESOURCES FROM CB INSIGHTS:

• State of Venture Q3’25 Report
• State of Fintech Q3’25 Report
• State of AI Q2’25 Report
• State of Digital Health Q2’25 Report
• State of Tech Exits H1’25

The post State of AI Q3’25 Report appeared first on CB Insights Research.

As global competition intensifies and supply chains fracture, the ability to deploy capital into technologies that make the US more self-reliant, adaptive, and secure has become a national priority. 

JPMorgan Chase’s initiative is designed to meet that moment, representing both a financing bonanza for US-based companies across the 27 strategic sub-areas identified by the banking giant and a capital deployment challenge. 

To help, we’ve used CB Insights’ predictive signals, including our outlook score for company success, Mosaic, to shortlist 79 companies for JPMorgan Chase to consider investing in. 

We map them all below, categorized across all but 4 of the strategic areas identified by JPMorgan Chase due to a lack of promising startups matching our criteria — full methodology available at the end of this report. Key recommendations follow.

Find our success and exit predictions for the full list of featured companies here

Key recommendations

Invest in AI as the backbone of manufacturing sovereignty 

A new class of industrial AI startups is aiming to rebuild the nation’s productive capacity across manufacturing, materials, and mobility. 

Companies like Skild AI (robot foundation models), Charge Robotics (factory automation), and Cartken (autonomous mobile robots) are reducing the cost gap for onshore production by automating physical work and logistics. 

Meanwhile, Earth AI and Periodic Labs are applying AI to accelerate materials discovery and secure domestic access to critical minerals and battery inputs. Material development platforms have recently captured investor attention as AI and quantum computing unlock new capabilities, becoming one of the hottest emerging markets (i.e., with a Commercial Maturity median of 3 or below) in the manufacturing sector. 

Source: CB Insights Markets Index (as of 9/9/2025)

Even legacy-heavy domains like shipbuilding are seeing AI-enabled autonomy from firms such as Seasats and Blue Water Autonomy, which aim to reduce costs in a historically high-cost, labor-intensive industry. 

Together, these companies are forming the digital-industrial layer for intelligent reshoring, using AI to make domestic manufacturing faster, adaptive, and globally competitive again.

Bet on autonomous defense systems to redefine deterrence

Modern conflict is evolving faster than traditional defense procurement can adapt, primarily driven by the proliferation of drone warfare with the invasion of Ukraine. A new generation of startups is filling the gap with software-defined, low-cost autonomy. 

Companies like Fortem Technologies (drone detection and interception) and Allen Control Systems (Counter-unmanned systems) are deploying scalable unmanned systems across air and land. Paired with Picogrid’s defense connectivity networks and Rune Technologies’ predictive logistics AI, these platforms create adaptive, sensor-rich meshes that can respond to threats in real time. 

The result is a rapidly emerging autonomous defense ecosystem that strengthens deterrence through speed, intelligence, and distributed resilience.

See the hottest defense tech markets ranked by equity raised over the past year here

Smarter and cleaner energy production and distribution will help maintain AI lead

The AI boom has made energy an ever greater strategic resource, making diversified, digitally managed generation essential to both national security and computational progress. 

Investing in startups focused on nuclear innovation, solar automation, and AI-driven grid resilience will help build a smarter, sovereign energy base. 

Companies like Pacific Fusion and Realta Fusion are redefining what clean baseload power can look like, while Aalo Atomics and Natura Resources advance modular fission systems that can be rapidly deployed near data and industrial centers. 

Paired with Antora and Rondo Energy’s storage systems and David Energy’s automated grid orchestration, these technologies form an adaptive, distributed energy network capable of self-balancing demand spikes from AI compute and industrial electrification. 

See the most promising startups helping manage demand spikes from AI here

Investing in this ecosystem builds not only clean power capacity but also strategic autonomy, ensuring the intelligence revolution runs on domestically controlled, resilient, and low-cost energy.

Support edge intelligence to make AI truly ubiquitous

The frontier of artificial intelligence is shifting from centralized data centers to the physical world, where latency, cost, and privacy demand on-device autonomy. 

Startups like EnCharge AI (compute-in-memory chips) and MemryX (edge inference for automotive and robotics) are re-architecting silicon for real-time decision-making, while Modal and Fireworks AI deliver serverless deployment layers that make model execution as elastic as cloud functions. 

On the software side, Together AI (an AI 100 2025 winner) and Liquid AI are advancing compact, efficient small language models that can run locally, enabling intelligence in drones, sensors, and industrial systems. 

Combined, these technologies are creating a distributed AI fabric that embeds cognition into every node of the economy, from factory floors to vehicles and handhelds. The result: lower latency, greater resilience, and unprecedented reach, as intelligence moves from “in the cloud” to everywhere work gets done.

Track emerging genAI startups, shifting markets, and breakout opportunities in real-time here

Finance the intelligent age security: AI, data, and infrastructure

As AI becomes the backbone of critical systems, the attack surface expands exponentially. 

A new cohort of startups is fortifying this frontier by integrating AI-native security, quantum-safe encryption, and infrastructure hardening into the fabric of the intelligent economy. 

Companies like TXOne Networks and Xage Security are protecting industrial and energy assets with zero-trust architectures built for operational technology, while TrustLogix and Concentric AI safeguard sensitive enterprise data through granular policy enforcement and autonomous monitoring. 

TrustLogix is among the AI security startups most likely to be acquired next, according to CB Insights Predictive Intelligence, as large cyber players have been on a M&A spree to seize this opportunity and integrate AI security features into existing offerings.

See the top AI security startups ranked by their M&A probability

Quantum Xchange ensures secure data transmission with quantum-resistant encryption, and HiddenLayer defends AI models themselves against poisoning, inversion, and adversarial attacks.

Together, these technologies create a secure AI infrastructure stack that spans digital, data, and physical domains, ensuring that as intelligence scales, trust scales with it. Investing in this layer isn’t just about cybersecurity; it’s about protecting the nervous system of the modern economy.

Methodology

We used CB Insights’ predictive intelligence to analyze thousands of startups and select the most promising ones that align with JPMorgan Chase Security and Resiliency Initiative’s 4 key areas and 27 sub-areas.

We looked across our 1,500+ Markets to identify the hottest areas based on equity funding raised over the past year and selected companies with the highest Mosaic scores (min 550) operating in these markets. We focused on private, early to mid-stage (Series B max), US-based startups. Data is as of 10/23/2025.

For information on reprint rights or other inquiries, please contact reprints@cbinsights.com.

The post 75+ investment targets for JPMorgan Chase’s Security and Resiliency Initiative appeared first on CB Insights Research.

This cohort cuts through the noise to spotlight 45 companies we expect to have an outsized, lasting impact over the next 5-10 years — from AI infrastructure that powers autonomous enterprise systems to vertical AI applications in healthcare, financial services, and manufacturing that solve operational problems.

Using CB Insights’ proprietary data — including Commercial Maturity, Mosaic, patents, business relationships, and funding — we identified 45 emerging players most likely to have a strong exit in the next 5-10 years.

See every startup in the Future Tech Hotshots 2025 Expert Collection

Key takeaways

• Agent infrastructure is the new frontier. The cohort reveals a decisive bet on agentic AI, with startups like Coval (AI agent testing), Questflow (multi-agent orchestration), and Syncari (agentic master data management) building the foundational tools that enable autonomous AI to operate reliably at scale. These companies are positioned for outsized impact because they’re creating the critical layer to embed AI into workflows — just as cloud infrastructure enabled SaaS, agent infrastructure will enable the next wave of autonomous enterprise software.

• The 45 hotshots have collectively formed over 110 business relationships since 2024. LLM data preparation company LlamaIndex leads the pack (18 partnerships), having partnered with incumbents like Microsoft and Databricks, while blockchain infrastructure API startup Crossmint has forged partnerships with Visa (to enable AI-driven on-chain payments) and Moneygram (to power new stablecoin cross-border payment experience). As these startups scale over the next 5–10 years, this early validation with enterprise incumbents will become harder to displace as customers build workflows around their products.

• Industrial AI is the most promising area, with companies in this space having experienced the highest Mosaic score increase over the last 6 months. This includes GIS platform Felt (+71 points in 6 months) and humanoid developer Persona AI (+57). This momentum reflects investor and customer recognition that industrial AI creates defensible moats through domain-specific datasets that take years to build. Unlike horizontal tools, this vertical expertise can’t be easily replicated, positioning these companies as prime acquisition targets for industrial incumbents seeking AI capabilities over the next 5-10 years.

• Elite management teams cluster in enterprise infrastructure. Top Management Mosaic scores concentrate in enterprise tech, with Lineaje (962/1000; software supply chain security platform), Maven AGI (956/1000; customer service AI agents), ProRata.ai (950/1000; AI-powered search and advertising), and Harmonic (876/1000; mathematical superintelligence) all led by executives hailing from incumbents like Google, Robinhood, and Stripe. These companies signal that the most experienced founders see enterprise infrastructure — not verticalized or consumer AI — as the category where technical depth and execution create the most competitive advantage.

Methodology

We used CB Insights data to analyze hundreds of VC-backed private tech companies with Mosaic scores of 600+ and an early commercial maturity score. 

Our scoring model factors in signals like investor quality, business relationships, Mosaic scores, key people data, and patents. We excluded companies with fewer than 100 employees. Data is as of 9/29/2025.

For information on reprint rights or other inquiries, please contact reprints@cbinsights.com. 

The post Future Tech Hotshots 2025: 45 emerging tech startups poised to make an outsized impact appeared first on CB Insights Research.

AI-powered coding has become one of the fastest-growing enterprise use cases for LLMs. Startups like Anysphere (maker of Cursor), Replit, and Lovable have crossed $100M in ARR in record time, with adoption surging across professionals and amateurs, small companies and enterprises alike.

Satya Nadella says that as much as 30% of Microsoft’s code is now written by AI. But this breakneck adoption is exposing new pain points for enterprises: volatile costs and emerging security risks as AI-generated code volumes surge.

As agentic AI transforms software development, tools that deliver shipping speed, cost certainty, and de-risk AI outputs will separate sustainable enterprise adoption from expensive experiments.

Using CB Insights’ predictive intelligence on private companies, we mapped 90+ companies offering AI-powered solutions across 8 software development markets to reveal how agentic AI is collapsing the development stack, identify which capabilities set winners apart from losers, and uncover which features matter most to enterprises.

Note: Our market map includes private AI startups operating in the software development space that have received funding in the last 2 years. Companies may operate in multiple markets. This market map is not exhaustive of companies operating in the space.

Explore these companies sorted by their Mosaic score here

Key takeaways

Agentic AI is collapsing the software development stack — orchestration will separate winners from losers

The future is agentic. End-to-end agents are not just speeding up coding, they are combining planning, implementation, testing, and deployment into unified workflows, pushing the market from many fragmented tools toward fewer intelligent platforms. Adopters are feeling the same way.

3 out of 4 companies offering solutions for the software development lifecycle (SDLC) now have semi- or fully autonomous agentic features, including multi-agent capabilities. Companies offering agentic features also average a Mosaic score of 653, a 100 points higher than the average for companies without agentic features, signaling stronger overall health and growth potential.

Agentic AI is expediting code creation and enhancing developer workflows better than traditional coding tools, but reliability issues and enterprise requirements, such as compliance, governance, security, and quality assurance, will continue to drive demand for specialized agents at each SDLC phase.

Winners will provide unified orchestration layers that integrate specialized, best-in-class agents for each software development phase with governance and human oversight built in.

To operationalize AI, enterprises should evaluate vendors on integration with existing development infrastructure, governance controls that enable oversight and auditability, and observability into agent performance and associated costs.

Inference cost management is emerging as a key feature as AI coding bills soar

Reasoning models have inflated output token volume by 20x, pushing vendors away from seat-based pricing toward usage-based and hybrid models.

In response, cost control solutions are emerging as critical infrastructure. Larridin tracks AI costs and measures tool effectiveness, Cline is building cost controls directly into its offerings, while Paid reached a $100M valuation, pioneering outcome-based billing.

Enterprises should prioritize vendors with built-in cost governance. Look for solutions that offer cost-monitoring dashboards, usage forecasting, and budget controls as token volumes surge.

As costs shift from predictable seats to variable usage, expand AI governance beyond engineering to include stakeholders across finance, operations, and marketing. Accessible vibe coding tools are democratizing software development, enabling these non-technical teams to become builders and making cross-functional cost oversight essential.

For AI adoption, enterprises should look beyond traditional billing models. Consider vendors like Paid, Alguna, or Orb which offer results-based, AI-native billing to align costs with business value rather than infrastructure consumption.

Enterprise selects agentic AI providers based on security, integration depth, ability to ship software faster, and value-based pricing

End-to-end agents demonstrate strong market validation with an average Mosaic score of 745, indicating robust health, reliability, and growth potential.

We analyzed buyer interviews from the highest-valued agentic AI providers to identify enterprise selection criteria:

• Security compliance for AI-generated code 
• Ease of use for non-technical users and the ability to ship end-to-end solutions faster, 
• Deep and seamless integration with existing systems
• Transparent pricing justified by measurable value

As enterprises scale AI adoption, prioritizing vendors with native integrations and robust AI code security is key to reducing implementation friction. Check our AI security market to identify vendors leading in security and derisking AI.

Beyond immediate selection criteria, specialized coding small language models (SLMs) are an emerging trend to watch. Relace.ai’s recent $23M funding signals SLMs’ ability to improve coding efficiency at lower costs, a key enabler for value-based pricing models.

The blended AI coding market is temporary — expect separate markets for consumer and enterprise

There is a democratization of AI developer tools. Non-technical and professional use cases are hard to distinguish and startups are currently willing to blend the 2. The total addressable market (TAM) is no longer one or the other as companies can sell into both.

While vibe and enterprise coding markets are currently intertwined, we expect two distinct markets to emerge, targeting individual vibe-coders and professional developers, respectively. For example, Replit recently switched from targeting professional developers to non-technical users and gained momentum as a result.

Companies that can successfully pivot or maintain dual product lines will be better positioned. This suggests looking for teams with the flexibility to adapt their positioning as the market matures.

With an enlarged TAM, this offers investors a chance to diversify portfolios across both enterprise-focused and vibe-code-oriented AI coding tools.

Early traction metrics can mask user composition risks. Investors should verify if startups are winning with their intended customer by checking out the vibe-coding startups with the highest Mosaic scores.

Market descriptions

Software design tools

The software design market provides tools that create blueprints, technical specifications, and user experience designs for applications before development begins. Key activities include high-level design (system architecture and UI/UX structure), low-level design (data models, database schemas, API specifications, and component interfaces), and interface design (wireframes, mockups, prototypes, and design systems). Some platforms use AI to accelerate diagram generation, architectural planning, and technical documentation. These tools enable teams to visualize and validate both system architecture and user experiences before writing code.

Code generation & assistance

This code generation and assistance market offers tools and platforms that use AI to generate code, provide autocomplete suggestions, explain code snippets, debug issues, and assist with writing functions or modules. Solutions are delivered as IDE integrations, native IDE features, standalone applications, or self-hosted platforms. They accelerate coding by reducing boilerplate, catching errors, and offering intelligent recommendations. Some platforms include partial agentic actions that automate routine coding tasks. While some companies extend into adjacent phases like testing or deployment, their core offering focuses on code creation and maintenance.

Code review

The code review market provides tools that facilitate the systematic examination of code by developers or peers to identify errors, bugs, vulnerabilities, and adherence to coding standards. These platforms automate code analysis, provide static analysis capabilities, and offer collaboration features for reviewing and discussing code changes. Solutions range from standalone review platforms to integrations within version control systems and CI/CD pipelines. They help teams enforce consistency and quality across codebases and catch issues before they reach production. Some vendors incorporate AI to suggest improvements, detect security flaws, or automatically flag non-compliant code patterns.

Test automation software

The test automation software market provides platforms and tools that automate software testing processes across web, mobile, API, and desktop applications. Solutions enable QA teams and developers to design, create, execute, and manage automated tests through scripted frameworks, low-code/no-code interfaces, and AI-powered test generation. AI capabilities include automatic test case generation, failure prediction, self-healing when applications change, and test suite optimization. These platforms integrate with CI/CD pipelines, version control systems, and development workflows to enable continuous testing throughout the software delivery lifecycle. Both traditional and AI-driven solutions reduce manual testing effort and increase test coverage through automated validation.

Release automation software

The release automation software market provides platforms that automate and orchestrate the deployment of software releases from development to production environments. These solutions manage release planning, deployment automation, environment coordination, and progressive delivery strategies including feature flags and rollback capabilities. Platforms enable teams to schedule releases, coordinate deployments across multiple environments, manage dependencies, and ensure controlled rollouts to production. Solutions integrate with CI/CD pipelines, version control systems, and monitoring tools to provide end-to-end visibility into the release process. The market serves software development and operations teams seeking to standardize release processes and reduce deployment risk.

Code documentation & knowledge management

The code documentation and knowledge management market provides platforms that help development teams create, maintain, and preserve technical knowledge throughout the software development lifecycle. Solutions automatically generate documentation from code, maintain technical wikis, create API references, and answer questions about codebases using AI. These platforms capture what code does and why decisions were made, preserving organizational knowledge as team composition changes. The market serves software development teams seeking to reduce onboarding time, and maintain up-to-date documentation.

Code modernization & technical debt management

Code modernization and technical debt management market encompasses solutions that help organizations refactor, upgrade, and maintain legacy codebases while systematically addressing accumulated technical debt. It includes tools and platforms that automatically analyze and assess legacy code quality, dependencies, and technical debt, migrate applications from legacy languages to modern frameworks and automate code translation, modernization, and cloud migration efforts.

End-to-end software development AI agents

The end-to-end software development AI agents market offers AI agents that function as virtual software developers and engineers, independently handling end-to-end and full-stack development tasks with zero to minimal human supervision. Agentic AI understands requirements, writes code, debug, test, and deploy – operating across the entire software development lifecycle (SDLC) autonomously.

The post The AI software development market map appeared first on CB Insights Research.

The surge was fueled by outsized mega-rounds to new decacorns — companies with $10B+ valuations — and the continued dominance of AI, which accounted for 51% of all funding and 22% of deals in Q3’25.

However, funding growth was far from uniform across sectors. Retail and healthcare saw quarterly declines, while fintech remained flat. The data suggests that investors are pulling back from traditional industries and doubling down on emerging technologies, especially AI.

Below, we break down the top stories from this quarter’s report, including:

• Funding surpassed $90B for the 4th consecutive quarter
• AI is on track to capture over 50% of total annual venture funding for the first time 
• Decacorns raise record funding, as quarterly tech mega-rounds reach a new high
• Humanoid robots captured the most deals for the 2nd quarter in a row
• Exits are rebounding despite companies staying private longer

Let’s dive in.

Download the full report to access comprehensive data and charts on the evolving state of venture across sectors, geographies, and more.

Top stories in Q3’25

1. Funding surpassed $90B for the 4th consecutive quarter

Venture funding exceeded $90B for the fourth consecutive quarter, reaching $95.6B in Q3’25. The year-to-date total surpassed $310B, marking the highest annual figure since 2022.

Deal count, however, fell to its lowest point since Q4’16, underscoring an ongoing trend: investors are writing bigger checks to fewer companies. This pattern has persisted for over a year.

AI maintained its stronghold on the venture market, capturing $47.8B in Q3, for 50% of total funding and 22% of deals. Both figures represent the second-highest quarterly levels on record, confirming AI as the primary driver of venture strength.

While the largest rounds of the quarter went to leading AI players, such as Anthropic and OpenAI, other standout fundraisers included:

• China Fusion Energy, which raised $1.6B to advance fusion reactor engineering and commercialization.
• DBG Health raised $1B to provide a range of healthcare services.
• PsiQuantum, a developer of photonic quantum computers, raised a $1B Series E round.

Investors are also fueling a resurgence in hard tech — particularly in aerospace, defense, and advanced computing: 

• Aerospace funding reached $14.1B through Q3’25 and is expected to reach $18.9B by year-end — surpassing its 2021 record by 20%. 
• Defense tech raised a record $13.7B, driving the emergence of a new military-startup complex. 
• Quantum computing tripled its previous annual funding record, reaching $3.7B.

The data points to a venture market in transition — one defined by larger checks, fewer deals, and a growing concentration of capital in AI and hard tech.

2. AI is on track to capture over 50% of total annual venture funding for the first time

AI companies are capturing a record share of funding and deals this year, at 51% of funding and 22% of deals. They also claimed 7 of the 10 largest rounds this quarter.

The US is proving especially dominant in AI, attracting 85% of total AI funding and 53% of deals in 2025. Four of the 7 largest rounds this quarter were based in the US: Anthropic, OpenAI, Databricks, and Figure.

Funding to AI-enabled companies is also taking a significant share of traditional sectors:

• Retail tech declined to $5.4B, its lowest quarter since Q3’24, with AI startups raising 36% of annual funding.
• Digital health fell to $4.5B, marking its weakest quarter since Q4’24, with AI startups representing 63% of the sector so far this year.
• Fintech remained flat at $10.9B quarterly, with AI firms accounting for 23% of total fintech funding in Q3’25, its 2nd highest quarter on record.

AI is creating a clear split in the venture ecosystem, with AI startups capturing an outsized share of capital and mega-rounds, while non-AI startups face tighter funding conditions.

The rapid rise in AI valuations raises questions about long-term sustainability, as many companies are priced for winner-take-all outcomes across categories, particularly in saturated markets like coding agents & copilots, where dozens of similar startups compete as margins tighten.

The current environment reflects a flight to quality. While AI continues to drive momentum and capital concentration, the market is gradually shifting toward fundamentals — where execution and efficiency, not just promise, will determine which companies justify their valuations.

3. Decacorns raise record funding, as quarterly tech mega-rounds reach a new high

The venture landscape is moving beyond unicorns to decacorns — companies valued at $10B or more. Decacorns raised a record $94.5B through Q3’25, surpassing the previous record of $46.3B in 2024.

However, the number of decacorn deals is almost half as much as it was in 2021 when it reached $45.5B — from 60 deals to 32 this year — revealing the high funding concentration among the very largest companies — primarily leading AI startups.

AI leaders raising at decacorn valuations include developers xAI, Scale, and Perplexity, defense startups Anduril and Helsing, and fintech company Ramp.

Beyond decacorns, $100M+ mega-rounds for tech companies also hit record levels. September saw 52 tech mega-rounds in total, with 70% of capital allocated to companies focused on making AI infrastructure more affordable at scale.

Many AI infrastructure companies that raised mega-rounds in Q3’25 have already generated substantial revenue. Invisible Technologies reached $134M in 2024, Baseten reportedly grew 10x YoY, while Rebellions projected $72M in revenue. This shift separates real businesses from overvalued concepts as scrutiny intensifies.

Decacorns and mega-rounds are defining the current venture landscape. The market is bifurcating not only between AI companies and the rest, but also between decacorns and mega-round recipients vs. everyone else.

We expect the gap between well-funded companies and the rest of the venture ecosystem to continue widening as capital concentrates among market leaders who are building critical infrastructure and enterprise solutions.

4. Humanoid robots captured the most deals for the 2nd quarter in a row

AI markets dominated the most active deals in Q3’25, including AI-powered humanoids, AI software applications, and autonomous driving. 

Industrial humanoid robots captured 17 deals — more than any other market — continuing momentum from Q2’25, when it also led with 23 deals. New humanoid robot unicorns also emerged — Zhiyuan Robot and Unitree Robotics — bringing the total to 4.

Humanoid deal activity extended outside of the industrial sector in Q3. Healthcare humanoid robots secured 7 deals, ranking just outside of the top 10 markets. Figure led both the industrial and healthcare humanoid markets, raising a $1B Series C round at a $39B valuation, making it the 9th most valuable private company globally.

Investor interest in humanoid robots is driven partly by physical AI enabling new robotics capabilities, giving humanoids commercial promise that was not previously possible.

But despite deal activity and future potential, humanoids remain years away from widespread deployment. Developers still face fundamental challenges with inference, dexterity, reliability, and cost, which limit initial use cases to structured environments like factories and warehouses with a controlled and predictable set of tasks.

Autonomous driving showed particular strength among markets powered by physical AI. Both autonomous trucking systems and autonomous driving systems captured 8 deals each, ranking among the most active markets by deal count, alongside prominent AI categories such as coding AI agents, AI agent development platforms, and LLM developers.

5. Exits are rebounding despite companies staying private longer

Exits are recovering, but the numbers also reveal a fundamental shift in how long startups remain private before going public or getting acquired.

M&A and IPO activity both rebounded in Q3’25, partly driven by maturing AI startups that created more exit opportunities. M&A deals rose 8% from last quarter to 2,324 — the highest total since Q3’22. AI M&A activity remained elevated at 172 deals, contributing to the increase.

Fintech M&A contributed heavily to the rebound, rising to 249 deals — its highest level since Q1’22. Healthcare M&A also hit its strongest level since Q1’23, with 3 of the top 10 M&A transactions going to healthcare companies.

IPO activity climbed 45% from 95 to 138 — the highest quarterly total since Q3’23. AI and fintech contributed to the uptick, but software companies dominated the largest offerings. The biggest IPOs went to Figma and Klarna. The only hardware exception was China-based Best Semi, a semiconductor equipment manufacturer.

The Q3 exit rebound reflects improving conditions and suggests a broader recovery ahead, especially if interest rates continue to decline.

Despite increased exit activity, companies are staying private longer, with the time to exit rising from 12.2 years in 2015 to 15.9 years in 2025.

The ability to raise at decacorn valuations while staying private removes the pressure to go public for capital. Companies can now scale to a massive size, hire top talent through liquid secondary markets, and maintain founder control — all without the quarterly earnings pressure or regulatory burdens associated with going public.

Exit levels are recovering, suggesting that the market is normalizing, but the structural shift toward longer private tenures is likely to remain. The venture lifecycle is undergoing a fundamental change, with companies now possessing viable paths to scale privately that did not exist a decade ago.

If you aren’t already a client, sign up for a free trial to learn more about our platform.

RELATED RESOURCES FROM CB INSIGHTS:

• State of Venture Q2’25 Report
• State of Fintech Q3’25 Report
• State of AI Q2’25 Report
• State of Digital Health Q2’25 Report
• State of Insurtech Q2’25 Report
• State of Tech Exits H1’25

For information on reprint rights or other inquiries, please contact reprints@cbinsights.com.

The post State of Venture Q3’25 Report appeared first on CB Insights Research.

Rohit Verma, President & Chief Executive Officer of Crawford & Company, shares executive insights on insurance affordability.

A few months ago, I sat down to review my monthly expenses and was stunned. My auto insurance premium had climbed over 30%, and my home insurance has almost doubled since the start of the pandemic. These aren’t discretionary expenditures — they’re essential protections. Friends and colleagues have shared similar stories, some even seeing insurance costs exceed their monthly mortgage payments, a shift unimaginable just a few years ago. One friend recently asked me, “Where is this going? Are we heading toward a future where insurance becomes unaffordable for the average household?”

In recent years, insurance rates have increased considerably, largely due to severe property losses from wildfires, storms, and hurricanes. Social inflation is also straining the industry. These pressures have forced carriers to raise deductibles and prices; trends reflected in property loss ratios. Necessary to maintain industry stability, these actions have also burdened policyholders, challenging affordability and accessibility, and intensifying financial strain. Crawford adjusters witness how today’s risk environment challenges insurance’s core purpose: transferring risk from insured to insurer.

Thankfully, recent reinsurance renewals suggest relief may be on the horizon, barring major catastrophes. As these dynamics unfold, we expect pricing and deductibles to ease, leading to normalization in the next 12 to 18 months. However, current premiums aren’t sustainable, and the industry must act now to support customers through this period.

This short-term relief opens a critical window to implement more sustainable solutions, like enhanced technology adoption, resilient rebuild strategies, and stronger safeguards against legal exposure and fraud. Smarter technology use, resilient rebuilding, and proactive stances against damaging practices offer meaningful progress toward stability, though none are quick fixes.

Technology is a vital lever to alleviate cost pressures and improve the insurance experience as the market normalizes. Advanced solutions, such as agentic AI, predictive analytics, and AI-powered risk insights, are already transforming insurance. These technologies streamline workflows, enable faster, more accurate estimates, and help insurers proactively manage risks.

These innovations go beyond efficiency; they reshape the customer experience. Policyholders benefit from faster claims resolution, more transparent communication, and tailored risk management. Leveraging real-time data and automation, insurers deliver responsive, personalized service, restoring trust and confidence, even amid market pressures.

This report, developed with CB Insights and Crawford & Company, offers a data-driven roadmap for addressing affordability. By analyzing tech momentum and affordability across nine P&C lines, we pinpoint where technology most effectively reduces loss costs and benefits policyholders. Findings show targeted adoption of advanced tech, especially in cyber, homeowners, and auto, delivers the greatest impact.

I encourage all of us as industry stakeholders to use this insightful report to explore investments, partnerships, and solutions. Together, we can build a more resilient insurance ecosystem — keeping coverage accessible, affordable, and customer-centric for all.

Rohit Verma
President & Chief Executive Officer
Crawford & Company

Overview

Insurance coverage is becoming increasingly unaffordable for businesses and consumers, with loss costs rising rapidly due to factors like extreme weather, labor shortages, and supply chain disruptions.

To improve affordability, insurance companies must prioritize innovative ways to deploy technology across loss prevention efforts. While technology alone is not enough to improve affordability, it offers the most tangible opportunities to lower costs.

Below, we identify top tech-driven loss prevention opportunities to improve insurance affordability across nine P&C insurance lines of business. We rank these opportunities across two axes:

• Tech momentum assesses startup ecosystem strength and tech applicability across a line of business. We measure tech momentum using CB Insights’ datasets such as deal activity, company headcount, and our proprietary Commercial Maturity — which measures a private company’s ability to compete or partner — and Mosaic scores — which measure the overall health and growth potential of private companies.
• Affordability pressure assesses the impact of loss cost increases for policyholders across a line of business. We evaluate affordability pressure by surveying Crawford & Company’s global claims experts, coupled with CB Insights’ Public Company Financials data.

Key takeaways: Opportunities with the greatest potential impact

1. Cyber leads in startup momentum, offering insurers the richest landscape of tech innovation to improve affordability.
2. Homeowners’ insurance faces the greatest affordability pressure, making it the most urgent line for loss-prevention technologies despite limited ready-made solutions.
3. Commercial and personal auto face an innovation gap, with high affordability pressure but low startup momentum — requiring insurers to carefully vet and selectively scale emerging solutions.

1. Cyber leads in startup momentum, offering insurers the richest landscape of tech innovation to improve affordability.

Cyber insurance faces only moderate affordability pressure, primarily from data breaches and hacks, compared to other lines like auto and homeowners that are more exposed to medical costs, materials shortages, and natural catastrophes (NatCats). What makes cyber unique is that leading players such as Coalition (an insurtech unicorn with $300M+ in revenue in 2024) actively embrace technology to prevent and reduce losses at scale.

Because the line is inherently tech-centered, nearly every new entrant presents a potential opportunity to lower loss costs. Risk management practices within cyber have generally reduced claims severity, as noted by Greg Smith, President, Canada at Crawford & Company:

“A larger number of smaller cyber claims are reducing severity, as is a more disciplined focus on underwriting and loss control in this space. Loss ratios for cyber lines in Canada have improved significantly in recent years because of improved underwriting and risk management.”

Between July 2022 and June 2025, cyber-focused startups completed more than 2,500 deals of at least $100K in funding, more than any other line of business analyzed. With a median CB Insights’ Mosaic score (success probability) of 615 out of 1,000, cyber startups tie general liability for the highest score across the 9 lines of business.

This surge of new entrants reflects rising enterprise concerns around AI security, which has spiked in executive commentary since the release of ChatGPT in late 2022. For example, Knostic, founded in 2023 and recently doubling its headcount from 22 to 44, helps enterprises identify and mitigate large language model data leakage risks.

Looking ahead, as individuals and businesses increasingly adopt AI, advanced cyber risk detection and rapid response capabilities will become critical. Insurance companies should prioritize evaluating partnerships with cyber startups or building comparable in-house capabilities to pass affordability gains onto policyholders.

2. Homeowners’ insurance faces the greatest affordability pressure, making it the most urgent line for loss-prevention technologies despite limited ready-made solutions.

Homeowners’ insurance experienced the steepest estimated loss ratio increases among all business lines analyzed, coinciding with rate increases spurred by natural catastrophes (NatCats). Survey respondents also pointed to fraud, labor shortages, and materials shortages as amplifying factors.

Reducing loss costs within homeowners’ insurance will depend on consistent data availability for individual homes and surrounding communities, as noted by Tim Butler, Head of Contractor Connection & CRD, Australia at Crawford & Company:

“For quite a number of years, there has been consistent talk of preventative measures in the form of water pressure meters and standard home tech. Unfortunately, it appears adoption of this technology remains an uphill effort. There is, however, an increase in the use of data available, which appears to be creating more consistency around loss cost.”

While more than 1,900 startup deals touch the homeowners’ space, many of these products were not designed specifically for insurers, requiring carriers to actively identify and adapt external technologies for loss prevention.

For example, Atmosic is developing low-power Internet of Things (IoT) charging infrastructure that insurers could implement via homeowner-provided sensors. Homeowners could receive these sensors at the start of hurricane season, ensuring extended power to generate data on risks that could result in a costly claim.

Beyond sensors, insurance companies should also eye non-traditional data and risk engineering methods to improve affordability. For instance, Figure, a humanoid robotics company founded in 2022 and backed by Bezos Expeditions, Microsoft, NVIDIA, and OpenAI, is targeting household deployments. In the future, humanoid robots could proactively maintain homes, generate maintenance data, and provide insurers with differentiated insights to mitigate loss risks.

Looking ahead, affordability improvements in homeowners’ insurance will require a broad set of technologies that support both loss control and proactive risk management. NatCat events, in particular, will continue to stress the market, presenting increasingly pressing needs to reduce loss costs to the greatest extent possible.

3. Commercial and personal auto face an innovation gap, with high affordability pressure but low startup momentum — requiring insurers to carefully vet and selectively scale emerging solutions.

Commercial and personal auto rank lowest in tech momentum across all lines of business analyzed. Many of these startups are in the electric vehicle space, a sector now facing market headwinds with reduced executive attention and a pullback in dealmaking.

Despite weak innovation supply, both auto lines remain under high affordability pressure, second only to homeowners. Survey participants cited macro-economic factors such as rising materials costs and social inflation as primary contributors to worsening claim trends, with Steve Blakemore, Managing Director, U.S. Loss Adjusting at Crawford & Company, noting:

“Cost of materials and specialized repair processes to include aluminum bodies and e-vehicles have increased significantly beyond affordable deductibles.”

Technology opportunities for loss prevention exist, but they require disciplined evaluation — and need to extend beyond established telematics capabilities. For example, AtoB is a payments platform for the trucking industry with investors including Bloomberg Beta and Mastercard. The company’s revenue is projected to reach $100M by the end of 2025.

Coupled with existing data from telematics capabilities, insurance companies could utilize AtoB’s spending data to offer policyholders predictive notifications — for instance, guidance on routes to avoid areas with a higher likelihood of collisions.

Insurance companies should also identify opportunities related to autonomous vehicles, particularly after Waymo’s $5.6B Series C funding round in October 2024. Autonomous vehicle technology offers insurance companies potentially valuable data points that can inform loss prevention strategies as autonomous driving becomes more prevalent.

Looking ahead, improvements to auto insurance affordability through technology will require access to unique data sources that provide differentiated insights into driving risks. Carefully selected tech partnerships can provide insurance companies with access to this data, enabling them to offer policyholders proactive notifications that curb risky actions and prevent costly losses.

Line of business spotlights

Prioritize

Commercial property

Overview: Insurance coverage for business buildings, facilities, warehouses, and commercial real estate against property damage and operational risks.

Insurance Affordability Outlook placement:

• Affordability pressure ranking — 4th
• Tech momentum ranking — 5th

Data highlights:

• Pool of 800+ deals analyzed
  • Median deal size — $10.8M
  • Median current Mosaic score — 563 out of 1,000
  • Most frequently listed current Commercial Maturity — Scaling (level 4 out of 5)
• Survey respondents indicated that commercial property has significant exposure to weather-related natural catastrophes.
• Survey respondents also identified materials shortages and supply chain disruptions as highly relevant factors to commercial property.

Potential startup collaboration: Doss is an AI-enabled enterprise resource planning platform serving industries like construction, manufacturing, and supply chain. The company more than doubled its headcount between July 2024 and July 2025. An insurance company could pursue a partnership with Doss to gain data access to flag supply chain shortage risks that could otherwise lead to costly repairs and restorations.

Construction

Overview: Specialized insurance coverage for construction companies, contractors, and building projects, addressing construction-specific risks including accidents, defects, and site safety.

Insurance Affordability Outlook placement:

• Affordability pressure ranking — 5th
• Tech momentum ranking — 3rd

Data highlights:

• Pool of 1,100+ deals analyzed
  • Median deal size — $4.0M
  • Median current Mosaic score — 562 out of 1,000
  • Most frequently listed current Commercial Maturity Score — “Deploying” (level 3 out of 5)
• Construction has the third-highest tech momentum, indicating ample opportunity for innovation in the space.
• Survey participants indicated the following factors as highly relevant: weather-related natural catastrophes, labor shortages, materials shortages and supply chain disruptions.

Potential startup collaboration: AUAR builds mobile, robotics-powered micro-factories for home construction. The company partnered with industrial giant ABB in 2024 to expand operations in the United States. Given that micro-factories can reduce construction costs due to the potential need for less labor, materials, and transportation, insurance companies could offer AUAR-partnered construction companies less costly premiums.

Cyber

Overview: Insurance coverage against cybersecurity threats, data breaches, and digital risks affecting business operations and customer information.

Insurance Affordability Outlook placement:

• Affordability pressure ranking — 6th
• Tech momentum ranking — 1st

Data highlights:

• Pool of 2,500+ deals analyzed
  • Median deal size — $6.0M
  • Median current Mosaic score — 615 out of 1,000
  • Most frequently listed current Commercial Maturity Score — “Deploying” (level 3 out of 5)
• Cyber has the highest tech momentum, boosted above the others as the line is inherently tech-enabled. Cyber startups have the highest weighted-average score across deals analyzed in the report, and the largest deal count analyzed across the lines of business.

Potential startup collaboration: Lakera is a security platform for genAI applications. The startup has a CB Insights’ Mosaic score among the top 2% of companies globally, and is one of the world’s most-promising AI startups as a 2025 AI 100 winner. Insurance companies could partner with Lakera to offer the company’s tech to support policyholders’ AI agents, identifying and protecting against potential data breach attempts from malicious prompts.

Homeowners

Overview: Personal insurance coverage protects residential properties and personal belongings against property damage, natural disasters, and household risks.

Insurance Affordability Outlook placement:

• Affordability pressure ranking — 1st
• Tech momentum ranking — 7th

Data highlights:

• Pool of 1,900+ deals analyzed
  • Median deal size — $4.2M
  • Median current Mosaic score — 562 out of 1,000
  • Most frequently listed current Commercial Maturity Score — “Deploying” (level 3 out of 5)
• Homeowners’ insurance leads in affordability pressure, with analysis zeroing in on claims severity and loss ratio change (the greatest across lines analyzed) as key pain points.
• Survey respondents indicated that homeowners’ insurance has significant exposure to weather-related natural catastrophes.

Potential startup collaboration: Honey Homes is an on-demand maintenance service currently serving homeowners in California, Illinois, and Texas. Insurance companies could evaluate partnerships with Honey Homes to gain access to trending service requests in localized areas, like upticks in window replacements across older homes. Insurance companies could then derive data-driven signals from those requests to inform preventive action for potentially costly events, such as sending plywood and sandbags to homeowners in a projected hurricane path.

Vet

Commercial auto

Overview: Insurance coverage for vehicles used in business operations, including fleet management, trucking operations, and commercial transportation risks.

Insurance Affordability Outlook placement:

• Affordability pressure ranking — 3rd
• Tech momentum ranking — 9th

Data highlights:

• Pool of 500+ deals analyzed
  • Median deal size — $7.0M
  • Median current Mosaic score — 613 out of 1,000
  • Most frequently listed current Commercial Maturity Score — “Deploying” (level 3 out of 5)
• Commercial auto had the lowest tech momentum ranking due to below-average company scores and deal count.
• Materials shortages, medical costs, and social inflation are key factors in commercial auto claims.

Potential startup collaboration: Outpost offers a network of managed freight terminals, featuring a gate management platform that reviews truck data like license plates and registration numbers using computer vision technology. The company doubled its financial capacity in September 2025 to $1B. Insurance companies could pursue a partnership conversation with Outpost to gain access to gate data for risk modeling purposes.

General liability

Overview: Broad business insurance covering third-party claims for bodily injury, property damage, and operational risks arising from normal business activities.

Insurance Affordability Outlook placement:

• Affordability pressure ranking — 7th
• Tech momentum ranking — 6th

Data highlights:

• Pool of 200+ deals analyzed
  • Median deal size — $6.6M
  • Median current Mosaic score — 615 out of 1,000
  • Most frequently listed current Commercial Maturity Score — “Deploying” (level 3 out of 5)
• Survey responses indicated that general liability is experiencing increased claims severity. Social inflation and increased medical costs are key factors relevant to general liability claims.

Potential startup collaboration: Relay provides communication and location-tracking devices for frontline workers across industries, like entertainment, healthcare, and hospitality. The startup’s headcount has grown rapidly in recent years, with a projected revenue of $100M by 2027. Insurance companies could evaluate offering Relay’s product to business customers, like concert venues and restaurant operators. The tech deployment would support employee responses to potential claims-triggering risks, such as wet floors that could lead to slip and fall incidents.

Inland and ocean marine

Overview: Specialized coverage for goods in transit, commercial equipment, and property that moves between locations or operates in maritime environments.

Insurance Affordability Outlook placement:

• Affordability pressure ranking — 9th
• Tech momentum ranking — 2nd

Data highlights:

• Pool of 900+ deals analyzed
  • Median deal size — $5.3M
  • Median current Mosaic score — 607 out of 1,000
  • Most frequently listed current Commercial Maturity Score — “Deploying” (level 3 out of 5)
• Inland and ocean marine has the second-highest tech momentum, indicating strong opportunities to improve insurance affordability — despite the lowest pressure to improve affordability across all lines of business assessed.
• Survey respondents indicated the following factors as relevant: fraud, labor shortages, materials shortages and supply chain disruptions, and weather-related natural catastrophes.

Potential startup collaboration: Altana AI is a supply chain intelligence platform backed by Google Ventures and Salesforce Ventures, and — as a CB Insights 2024 Insurtech 50 winner — one of the world’s most-promising insurtech startups. The company offers products for business interruption risk and supply chain network planning. Insurance companies could explore Altana AI’s platform to gain visibility into potential supply chain risks and reroute shipments that otherwise face heightened risk for damage or loss.

Personal auto

Overview: Individual insurance coverage for personal vehicles, protecting against accidents, vehicle damage, and liability arising from personal driving.

Insurance Affordability Outlook placement:

• Affordability pressure ranking — 2nd
• Tech momentum ranking — 8th

Data highlights:

• Pool of 900+ deals analyzed
  • Median deal size — $7.3M
  • Median current Mosaic score — 606 out of 1,000
  • Most frequently listed current Commercial Maturity Score — “Deploying” (level 3 out of 5)
• Personal auto has the second-highest affordability pressure among lines of business analyzed, primarily due to loss ratio change estimates.
• Startups relevant to personal auto ranked the lowest in momentum due to a low weighted average score.
• Survey respondents indicated challenges around claims severity.

Potential startup collaboration: NoTraffic is a traffic management company that offers IoT devices to collect data at intersections and an AI platform to support traffic optimization decisions. NoTraffic is a NVIDIA partner and previously participated in the NVIDIA Inception Program. Insurance companies could evaluate partnerships with NoTraffic to support deployments that could reduce traffic accidents and lower claims costs.

Workers’ compensation

Overview: Mandatory insurance providing medical benefits and wage replacement for employees injured on the job, covering workplace accidents and occupational hazards.

Insurance Affordability Outlook placement:

• Affordability pressure ranking — 8th
• Tech momentum ranking — 4th

Data highlights:

• Pool of 100+ deals analyzed
  • Median deal size — $5.0M
  • Median current Mosaic score — 581 out of 1,000
  • Most frequently listed current Commercial Maturity Score — “Deploying” (level 3 out of 5)
• Survey participants generally viewed medical cost increases and social inflation as relevant to workers’ compensation insurance.

Potential startup collaboration: Protex AI offers a computer vision product that gives customers visibility into workplace risks (like speeding forklifts in warehouse environments) to prompt intervention by environment, health, and safety teams. DHL and Marks & Spencer are among Protex AI’s customers. Insurance companies could seek a partnership with Protex AI to offer the platform to policyholders facing elevated risks for workers’ compensation claims, like manufacturers and warehouse operators, due to workplace injuries.

Monitor

No lines of business fall within this category.

Methodology

The insurance affordability outlook provides an informational visual framework for insurance leaders to identify opportunities to improve insurance affordability for policyholders.

The opportunities encompass potential tech deployments for loss prevention and response across 9 lines of business: commercial auto, commercial property, construction, cyber, general liability, homeowners, inland and ocean marine, personal auto, and workers’ compensation.

The resulting visual plots the lines of business relative to one another across 3 categories:

1. Monitor — Lines of business with lower signals to warrant investment in tech to support insurance affordability. Insurance leaders should track developments in this space for future consideration.
2. Vet — Lines of business with moderate signals to warrant investment in tech to support insurance affordability. Insurance leaders should evaluate potential partnerships and tech deployment, pursuing the most-promising opportunities as innovation activities.
3. Prioritize — Lines of business with strong signals to warrant investment in tech to support insurance affordability. Insurance leaders should prioritize partnerships and tech deployments to operationalize.

Calculations across 2 axes — tech momentum and affordability pressure — guide plotting for each line of business.

Tech momentum

Tech momentum assesses startup ecosystem strength and tech applicability across the 9 lines of business.

This report leverages CB Insights’ AI-enabled deal search to identify 8,900+ venture-backed equity deals of at least $100K across the lines of business between July 1, 2022, and June 30, 2025. Deals were identified based on keywords specific to the lines of business, and some deals were excluded from the analysis. Deals analyzed in this report are not mutually exclusive, although the aggregate total constitutes approximately 9% of venture dealmaking between Q3’22 and Q2’25.

We generate a score for each deal that reflects the company’s momentum within the marketplace. The score uses CB Insights’ data, such as deal activity, company headcount, and proprietary Commercial Maturity and Mosaic scores. We include startups from across the venture landscape, although we assign greater weight to insurtechs given their direct relevance to the insurance market. Weighted average calculation guides the final ranking of the scores and the total number of deals analyzed across each line of business.

Affordability pressure

Affordability pressure evaluates the approximate impact of loss costs on insurance affordability for policyholders across the 9 lines of business.

We leverage 2 different data sources to measure affordability pressure for each opportunity:

• Survey of claims experts — 64 Crawford & Company employees provided directional insights across 160 line of business-specific responses on global loss trends.
• Loss ratio changes among publicly-traded insurers — Loss ratio changes between 2022 and 2024 across a sample of 15 companies: AIG, Allstate, American Financial Group, Arch Capital Group, AXIS Capital, Chubb, Everest Group, GEICO, Mercury General Corporation, Old Republic Insurance Group, Progressive, Selective Insurance, The Cincinnati Insurance Companies, Travelers, and W.R. Berkley. The data points are approximate based on the business mix of each of the 15 companies, and mindful of loss ratio dynamics in recent years due to the potential (or actual) impact of rate and price increases.

The final ranking is guided by the survey outputs and loss ratio change outputs, supported by ChatCBI reasoning leveraging data from across CB Insights’ Business Graph.

Additional notes

CB Insights has provided the information contained in this report for informational purposes only and does not constitute an endorsement or recommendation by CB Insights. Reasonable efforts have been made to ensure the accuracy of the information, and CB Insights makes no representation or warranty, express or implied, as to its completeness or accuracy.

Crawford & Company has not vetted, nor does it endorse, any of the companies or technologies mentioned in this report. These references are illustrative in nature and should be viewed solely as examples rather than recommendations.

The insurance affordability outlook is not an investment analysis, and should not be used to guide financial- or investment-focused decisions, including those pertaining to any insurance company operating across the analyzed lines of business. In addition, the report leverages a non-actuarial analysis and should not be used to discern financial performance (including loss ratio performance) across any lines of business analyzed.

This report is global in scope, although the analysis largely centers on the United States. Regulations and market dynamics differ across geographies, and the report does not account for every nuance across the industry.

We selected the 15 companies for the loss ratio change analysis due to comparable data reporting practices of loss ratio performance on annual reports using CB Insights’ Public Company Financials data. The report uses loss ratios as underlying loss costs are often not reported in a standardized format on annual reports. Loss ratios include loss adjustment expenses and typically spans lines of business. The changes in loss ratios were used to derive signals for lines of business subject to more affordability pressure.

The insurance affordability outlook is not absolute. The insurance industry and broader tech landscapes are subject to constant change, so future developments have the potential to impact the findings presented in this report.

About

Crawford & Company

Crawford & Company® is a leading global provider of quality claims management and outsourcing solutions with an expansive network of experts serving clients in more than 70 countries. Our unique ability to combine innovation and expertise advances our purpose to restore lives, businesses and communities across the globe. For over 80 years, clients have trusted Crawford to care for their customers as a seamless extension of their brand, keeping the focus where it belongs—on people. More information is available at www.crawco.com.

Contact: info@us.crawco.com

CB Insights

Headquartered in New York City, CB Insights is the leading provider of AI for market intelligence. The company aggregates, validates, and analyzes hard-to-find private and public company data. Its powerful AI tells users what it all means to them personally. The world’s smartest companies rely on CB Insights to focus on the right markets, stay ahead of competitors, and identify the right targets for sales, partnership, or acquisition. Visit www.cbinsights.com for more information.

Contact: researchanalyst@cbinsights.com

The post The insurance affordability outlook: Opportunities to alleviate insurance’s affordability problem with technology appeared first on CB Insights Research.

We’ve used generative AI, combined with our proprietary data on these companies and their markets, to create the following scouting reports — in just one click on CB Insights.

CB Insights customers can download the book using the sidebar and track all companies using the Expert Collection.

At the conference, be sure to:

• Visit the ITC Discovery Desk powered by CB Insights to speak with a strategist and receive full SWOT analyses to help determine who to meet and why.
• Attend the unveiling of the Insurtech 50 report, featuring the world’s most promising insurtech startups, on Thursday at 11am Pacific in Oceanside Ballroom B.
• Claim your 10 days of free predictive private market intelligence now.

The post Book of Scouting Reports: ITC Vegas 2025 appeared first on CB Insights Research.

In September, private companies globally raised 1,400+ early-stage rounds (noting this total will rise as more deals are published retroactively). Over 25% of startups that raised rounds are building AI-enabled products and services.

Download the full report to access comprehensive CB Insights data on early-stage activity, including top investors & deals, valuation data, and our predictive signals. Below, we highlight notable trends to watch.

Emerging trends & categories to watch

Click the links to see underlying deal activity. Categories are not mutually exclusive. 

Review the full deal dataset

AI agents

Similar to last month, companies targeting AI agent applications raised over 50 deals (out of 1,485). Key trends to note include: 

• “Smart Money” is all in on AI agents: The top 25 VCs identified by CB Insights backed 13 AI agent startups in September. This represents nearly 20% of all of the early-stage activity from these VCs in the month. Focuses include security (Akto, Fabrix Security, Terra Security) and governance, risk, and compliance (Geordie, Zania), indicating enterprise adoption and risk management are key investment priorities. 
• Customer service is one of the most established use cases but is still seeing early-stage traction: AI agents handling customer service, support tickets, and user interactions represent one of the largest early-stage agent categories in September (8+ deals). Support operations have clear unit economics, high volume repetitive tasks, and direct cost savings compared to human agents, driving continued activity here. Top companies to watch based on Mosaic scores include Doo (Mosaic: 747) and Rauda AI (Mosaic: 687). 
• Emerging voice AI sector: Voice and phone agents are attracting dedicated investment (6 deals, 11% of agent activity) as investors bet on solutions that can tackle the unique technical challenges of voice interactions (i.e., real-time latency requirements, natural speech processing, emotional intelligence, etc.). Confido Health and Prosper, for example, are focused on healthcare applications. Meanwhile, Vida and Vaani Research are building infrastructure to develop voice AI/phone agents. Review the voice AI development platforms market to compare 30+ vendors in the space.

Robotics

Companies building robots, and the systems that power them, raised over 70 deals in the month. 

Within robotics, defense & security applications led early-stage activity (17 deals, 24% of total robotics activity), reflecting geopolitical tensions driving investment in autonomous defense systems and surveillance.

Other notable traction is in foundation models and operating systems for robots, as investors bet on horizontal platforms (4 deals, 6%):

• The leading September round in this category went to DYNA, which raised a $120M Series A. The company is ranked as a “challenger” in our robot foundation model developers market. 
• Another highlight is InOrbit, which focuses on robot orchestration and was founded by Florian Pestoni and Julian Cerutti, who have experience at a range of large tech companies, including Google, Atlassian, Meta, and IBM. The founders’ strong pedigree is reflected in the company’s CB Insights Management strength score, which ranked the highest of all the companies raising early-stage deals in September.

Management strength score

CB Insights’ Management strength scores (out of 1,000) the founding and management team’s prior achievements and likelihood of achieving future success, like a high-value exit. 

Especially at the earliest stages of the startup lifecycle, the strength of the management team serves as a key signal of potential. 

AI for scientific discovery & materials development 

Three of the largest early-stage rounds of the quarter went to companies looking to accelerate scientific discovery and materials development with AI: 

• Periodic Labs ($300M Series A at $1B valuation)
• Lila Sciences ($235M Series A at $1.23B valuation)
• CuspAI ($100M Series A at $520M valuation) 

Both Periodic Labs and Lila Sciences are also building “autonomous labs” — with AI designing, conducting and iterating on experiments. All 3 companies are operating at Commercial Maturity level 2/5 (Validating), indicating they’re still testing and refining their products.

Methodology

This report includes equity early-stage financings (convertible note, angel, pre-seed, seed, Series A) to private companies in August 2025. We excluded companies that are later-stage that raised an angel round or convertible note in the month. Categorization based on company descriptions.

For information on reprint rights or other inquiries, please contact reprints@cbinsights.com.

The post Early-Stage Trends Report: Smart Money is all in on AI agents, the rise of autonomous labs, and more in September appeared first on CB Insights Research.