While OpenAI and Anthropic continue to dominate headlines, Q2 2026 data shows that AI investment is shifting toward a much more physical landscape: robotics, chips, and industrial automation.

Artificial intelligence remains, by far, the main driver of global venture capital, but it also reveals that the category gaining the most momentum is no longer confined to the digital world.

Physical AI: A Category Gaining Momentum

Cumulative investment in Physical AI has already reached $69.4 billion so far this year.

In addition, the United States currently accounts for 67% of all Physical AI investment, up from 51% just three years ago, highlighting a clear strengthening of its leadership in this segment.

Beyond the amount of capital invested, activity in the sector is also accelerating.

Industrial humanoid robotics developers led the analyzed categories by number of deals, while foundation models for robots also ranked among the most active verticals of the quarter.

A New Generation of Unicorns

The AI ecosystem continues to create companies of exceptional value.

During the quarter, 37 new AI unicorns emerged, the highest number since 2022.

Among them, DeepSeek stands out, reaching a valuation of $59.2 billion and establishing itself as one of the most relevant players in the global AI landscape.

Also standing out is Project Prometheus, a company specializing in AI applied to industry and backed by Jeff Bezos, which has already reached a valuation of $41 billion.

These figures show that the market continues to reward companies capable of developing differentiated technology with clear industrial applications.

Infrastructure Enters Its Own Growth Phase

The AI revolution is not limited to software.

Companies focused on hardware and infrastructure are also attracting significant attention from the markets.

One of the best examples is Cerebras, which went public with a valuation of nearly $40.6 billion, becoming one of the largest IPOs of the year for a company specializing in artificial intelligence chips.

All of this highlights that the development of AI depends not only on better models but also on the infrastructure capable of supporting them.

What Can Founders Learn?

Just two years ago, simply adding artificial intelligence to a product was enough to capture the attention of customers and investors.

Today, that phase appears to be over.

The market is rewarding companies capable of solving complex problems, building proprietary technology, and creating sustainable competitive advantages.

AI continues to attract billions of dollars, but the focus has shifted. An increasing share of investment is moving toward intelligent agents, infrastructure, and solutions capable of interacting with the physical world.

Everything points to the next major wave of innovation happening not only behind a screen. It will also reach factories, warehouses, hospitals, and any environment where artificial intelligence can transform real-world processes.

Source: State of Venture Q2’26, CB Insights