MetaProp is signaling a significant pivot in the investment landscape, moving away from the massive hype surrounding foundation models—the massive, general-purpose systems like GPT—toward what they call "Vertical AI." While 2023 was defined by speculative bets on the underlying technology, 2026 is seeing capital flow into specialized applications. These startups integrate deep industry expertise into sectors like real estate and insurance, where high barriers to entry and complex regulations create defensive "moats" that general models cannot easily cross.

The real value is now accruing in the application layer, which consists of tools designed to solve specific, high-stakes problems for enterprises. Examples include AI-native insurance platforms that can dynamically reprice risk in climate-volatile zones and intelligent building systems that automate energy consumption. These aren't just simple interfaces built on top of existing tech; they are sophisticated systems designed to replace legacy manual tasks, such as underwriting and claims adjudication, which have historically relied on inefficient spreadsheets and phone calls.

Investors are increasingly cautious about overpaying for the frontier models and hardware infrastructure where competition is most intense and profit margins are shrinking. Instead, they are looking for companies with proprietary data access that offers technology upside along with direct relevance to physical-world assets. This shift is accompanied by a record high in expected merger and acquisition activity, as larger corporations look to acquire AI specialists to modernize their operations and maintain a competitive edge in the global economy.

MetaProp is signaling a significant pivot in the investment landscape, moving away from the massive hype surrounding foundation models—the massive, general-purpose systems like GPT—toward what they call "Vertical AI." While 2023 was defined by speculative bets on the underlying technology, 2026 is seeing capital flow into specialized applications. These startups integrate deep industry expertise into sectors like real estate and insurance, where high barriers to entry and complex regulations create defensive "moats" that general models cannot easily cross.

The real value is now accruing in the application layer, which consists of tools designed to solve specific, high-stakes problems for enterprises. Examples include AI-native insurance platforms that can dynamically reprice risk in climate-volatile zones and intelligent building systems that automate energy consumption. These aren't just simple interfaces built on top of existing tech; they are sophisticated systems designed to replace legacy manual tasks, such as underwriting and claims adjudication, which have historically relied on inefficient spreadsheets and phone calls.

Investors are increasingly cautious about overpaying for the frontier models and hardware infrastructure where competition is most intense and profit margins are shrinking. Instead, they are looking for companies with proprietary data access that offers technology upside along with direct relevance to physical-world assets. This shift is accompanied by a record high in expected merger and acquisition activity, as larger corporations look to acquire AI specialists to modernize their operations and maintain a competitive edge in the global economy.