Venture Capital Intelligence Report
December 28, 2025 • Synthesizing insights from top-tier VCs
VCs are seeing a bifurcated market: AI infrastructure and defense tech commanding premium valuations while consumer and traditional SaaS face compression. Flight to quality continues as LPs demand cleaner metrics.
Series A/B rounds taking 20-30% longer to close. Seed still active but Series C+ requiring clear path to profitability. Corporate VCs stepping in more for strategic rounds.
AI infrastructure maintaining 15-25x revenue multiples while traditional software compress to 6-10x. Early-stage pre-revenue AI companies still getting premium treatment.
The picks-and-shovels play for AI gold rush. Hardware, specialized chips, and infrastructure software seeing massive demand as enterprises scale AI workloads.
Geopolitical tensions driving massive government spending. VCs betting on companies that can serve both commercial and defense markets with advanced technology.
Moving beyond general LLMs to specialized AI agents for specific workflows. Higher defensibility through domain expertise and proprietary data.
IRA funding catalyzing massive infrastructure build-out. Focus shifting from R&D to deployment and scale manufacturing of proven technologies.
As AI adoption accelerates, enterprises need new security frameworks. Massive opportunity in AI model security, data governance, and compliance.
The best defense tech companies will be AI-first from day one, not legacy contractors adding AI as feature
The biggest opportunities are where AI enables completely new business models, not just efficiency gains
Traditional databases and infrastructure weren't designed for AI workloads - complete reimagining needed
Government agencies building AI-first systems rather than retrofitting legacy infrastructure
New administration pushing AI adoption, DARPA increasing AI research funding, government efficiency mandates
$200B+ government IT spending could shift to AI-native solutions
Early signals from: Lux Capital, Founders Fund, 8VC
Companies to watch: Scale AI, Primer, Rebellion Defense
Multiple specialized AI models working together rather than single general-purpose LLMs
Cost optimization driving specialized models, better performance on domain-specific tasks
Could reduce AI inference costs by 10-100x while improving accuracy
Early signals from: Greylock, Kleiner Perkins, NEA
Companies to watch: Together AI, Mistral, Anthropic
Brain-inspired computing architectures for ultra-low power AI inference
Edge AI demands, sustainability concerns, breakthrough in chip design
Could enable AI in every IoT device - trillion device opportunity
Early signals from: Intel Capital, Samsung Ventures, Founders Fund
Companies to watch: BrainChip, SynSense, Innatera
Previous: Red hot during 2020-2021 with massive rounds → Now: Significantly cooled, limited new funding
User acquisition costs skyrocketing, iOS privacy changes, limited monetization options, platform dependency risks
What Changed: Post-iOS 14.5 reality hit hard, TikTok dominance makes new social extremely difficult
VCs Cautious: Benchmark, General Catalyst, Accel
Previous: Billions invested in 2021-2022 → Now: Minimal new investment, focus on utility
User adoption failed to materialize, tokenomics proved unsustainable, regulatory uncertainty
What Changed: Market realized speculation ≠ sustainable gaming economy
VCs Cautious: Most traditional VCs backing away
Previous: Dominant category for decade → Now: Compressed valuations, higher bars
Market saturation, AI threatening to automate many workflows, customer consolidation pressure
What Changed: AI making traditional software workflows obsolete faster than expected
VCs Cautious: Most growth-stage VCs
Data network effects and proprietary inference optimization are the only sustainable moats in AI
💡 Focus on creating unique datasets through user interactions rather than just model fine-tuning
— Sequoia Capital
Chief AI Officers are emerging as new buyer persona - different from traditional IT buyers
💡 Tailor sales process to CAOs who care more about business outcomes than technical specifications
— Bessemer Venture Partners
Show clear path to positive unit economics by Series B or expect significantly longer fundraising cycles
💡 Model out profitability scenarios early and optimize for capital efficiency from day one
— General Catalyst
EU AI Act compliance becoming table stakes for enterprise AI companies globally
💡 Build compliance frameworks into product architecture now rather than retrofitting later
— Accel
Deal volume down 35% YoY but average deal size up 20%. Clear bifurcation between AI/defense winners and everything else struggling
Series C • Lead: Amazon • Others: Google, Spark Capital
Shows continued mega-investment in foundation model companies despite commoditization concerns
Foundation ModelsSeries E • Lead: Founders Fund • Others: a16z, General Catalyst
Largest defense tech round signaling institutional acceptance of defense investing
Defense TechnologyTake-Private • Key investors: Accel, CapitalG, Kleiner Perkins
Traditional RPA being valued as legacy tech despite AI integration efforts
Open source AI will lose to proprietary models in enterprise
Most believe open source will commoditize AI
Reasoning: Enterprise customers will pay premium for guaranteed performance and support
Their Bet: Backing proprietary model companies like Anthropic heavily
Consumer AI products will emerge as breakout category in 2025
Consumer AI has failed to find product-market fit
Reasoning: Interface breakthroughs and cost reductions creating new opportunities
Their Bet: Seed investments in stealth consumer AI companies
First $100B AI infrastructure company emerges by 2026
HIGHAndreessen Horowitz • Timeframe: 2026
Implications: Would create massive new category of infrastructure investing and validate AI infrastructure as permanent category
50% of new enterprise software will be AI-native by end of 2025
MEDIUMSequoia Capital • Timeframe: 2025
Implications: Traditional SaaS companies without AI strategy will become uninvestable
Government becomes largest buyer of AI services, surpassing enterprise by 2027
SPECULATIVELux Capital • Timeframe: 2027
Implications: Would shift entire AI industry focus toward compliance and security over pure innovation
Leading indicator of AI infrastructure demand sustainability
Sustained 80%+ utilization indicates continued infrastructure investment
Dropping utilization suggests AI investment bubble deflating
Real adoption vs hype indicator for B2B AI companies
Accelerating enterprise deployment validates vertical AI thesis
Slow adoption suggests need for more infrastructure before applications
Could create massive new market for AI companies
Major federal AI contracts signal new mega-customer category
Procurement delays indicate regulatory/security obstacles