The Savvy Logic Behind VC Bets In ‘Uninvestable’ Sectors

By Thomas Cuvelier

Defense, energy, robotics and government have historically been classic no-go areas for VC investment. These “hard” industries have slow procurement cycles, tight regulatory oversight and high-friction customer migration in common. Legacy software vendors serving them have benefited from a barrier of complexity to innovate slowly without facing the risk of customer churn.

This made the victims of this year’s AI anxiety-driven sell-off all the more dramatic. Software juggernauts serving heavy industries — IBM, SAP, ServiceNow, Schneider Electric — have gone from safe bets to being the subject of investor scrutiny.

While headlines have attributed that sell-off to quick-fire Anthropic launches of tools for vertical industries, there’s more at play. The macro trend is a newfound founder enthusiasm to build AI-native entrants in legacy industries, and the backing they’re enjoying from VCs that can see the once-in-a-generation opportunity to disrupt entire industries.

Why investor perceptions are changing

Context is important. Geopolitical instability, supply chain pressure and energy security concerns have placed industrial resilience at the center of national policy.

Be it the U.S. or across Europe, policymakers are prioritizing investment in grid upgrades, transportation networks and public sector infrastructure, while also re-examining procurement and compliance systems that have slowed the adoption of emerging technologies that could bring said industrial resilience about quicker.

At the same time, quick advances in AI and agentic systems make it possible to build a new class of AI-native software tailored to “hard” industries through deep integration with verticalized tooling and specialist automation of critical workflows.

Age-old incumbent moats, like cumbersome migration cycles that put businesses off moving to new software providers, are also being challenged as embedded automation cuts migration processes down from weeks to days.

The creation of software in and of itself has become commoditised in the AI era, and more investors are spotting that operational depth, intuitive UI/UX, speed to market and seamless integration into complex real-world systems are traits of high-quality vertical software that startups are well-placed to build.

Investors are also realizing that most of the available value from horizontal SaaS has been extracted. In those early post-ChatGPT years, VCs widely backed AI companies building for non-regulated SMB adoption — exactly the audience that foundational model players like OpenAI and Anthropic are now making inroads with as they push into enterprises. Foundational models are general in nature, and their verticalization can therefore only stretch so far. Given this, AI-native products built for heavy industries are compelling and competitive propositions for VCs.

Growing faith that incumbents are vulnerable

There’s always been lots of skepticism among investors and tech executives that AI startups can meaningfully challenge incumbents that have been on top for decades. But those companies are operating over sprawling product architecture and processes that were built in the pre-AI era.

Pivoting from that state of affairs to AI-native systems is a massive undertaking, whereas new companies are being launched with those systems in place from day one. Incumbents also have a low incentive to innovate at pace when customer churn is limited. But in the current context of breakneck speed improvements to AI models and agentic systems, waiting for churn to show up will be too late.

Scepticism also risks overlooking the profile of outstanding founders building AI-native challengers. Some of the fastest-growing startups in defense, energy, government and the public sector are led by people who came directly from the same industries they are transforming. Their understanding of sector constraints and operational realities gives them an advantage over general software providers that lack the same specialism and experience.

Picking up pace

Savvy entrepreneurship and VC investors are colliding to make a play for hard sectors. Once seen as off-limits due to procurement complexity or regulatory burden, these sectors represent huge, untapped potential in the new AI-native era.

The emerging companies offering solutions designed for these industries with deep, vertical-specific tooling integration and critical workflow automation are well placed to command a growing share of overall AI funding as they serve customer pain points that have gone unanswered for years.

We are talking about disruption within markets worth trillions. The scale of the opportunity for growing VC interest in sectors they’ve historically avoided is no mystery or miscalculation. The vision is an ambitious one. Rather than simply building better software, the foundational sectors of the world economy are about to be reimagined.

Thomas Cuvelier is a partner for the U.S. and Europe at early-stage venture capital firm RTP Global. He currently oversees the deployment of the firm’s latest $1 billion fund, backing a range of AI-native startups building to disrupt legacy industries and business processes. In a personal capacity, Cuvelier wrote an angel check for Lovable at pre-seed.

Related Crunchbase queries:

• Global Physical AI Venture Funding In 2026
• Global Defense Tech Venture Funding In 2026

Illustration: Dom Guzman