As global investors have continued pouring capital into the nascent boom in artificial intelligence, concerns about the concentration of the U.S. equity market in the companies leading this trend—the so-called Magnificent Seven—have emerged across various sectors. On the private markets side, the bet on this technology is also prevalent, especially in segments such as venture capital. However, Hamilton Lane’s view is that these asset classes are more diversified than equities.
In the latest edition of its Market Overview, the private markets specialist emphasized that these seven stocks have been driving the performance of public equity portfolios for at least six years. Looking ahead, they expect this to continue for at least another two or three years.
As is often the case with major technological shifts, venture capital has been leading investments. “It is fair to say that the scale of the move into AI has been dramatic. More than 50% of the value of venture deals globally is now directed toward AI-focused businesses,” the report noted.
However, despite the strong allocation, these asset classes do not carry the same level of concentration risk as public markets, according to Hamilton Lane. The issue lies in the innovation that has been driving the AI stock market rally: large language models (LLMs).
AI beyond LLMs
“A significant part of the narrative for public market investors and the U.S. economy is the development of large language models to drive AI growth and adoption,” the firm stated in its report.
In this context, two scenarios are emerging on the horizon. Some market participants are betting on the realization of a new industrial revolution. Others do not see a scalable path for LLMs, with a ceiling on their development. “None of us knows the answer to this, but we all need to understand the questions in order to grasp what might come out of the AI box,” they noted.
This brings a key issue to the table: investing in AI—but how?
Currently, Hamilton Lane emphasized in its report, public markets are heavily anchored in LLMs. By contrast, private markets—including the VC ecosystem—cover a broader range of the AI spectrum and have more limited exposure to the scaling of these models. For example, they do not invest in data centers or the chips that are at the core of this technology.
“The venture world invests in the applications of LLMs. They invest in ‘things’ that make end LLM products work better, integrate better, and become easier to use,” the alternatives specialist noted.
According to the report, regardless of valuation concerns, the private market offers more diversified access to this technology than public markets.
For the firm, there is a 60% probability that we are not in an AI-related financial bubble. This means it is their base case, but the recommendation is clear: portfolios must be prepared for each scenario. This includes considering that public and private assets could behave differently if the promise of large language models does not materialize.
“Would anyone be surprised if, going forward, LLMs stagnate, but the applications using current LLMs thrive? That might not be a good environment for Magnificent Seven stock prices, but a group of venture-backed private companies would soar,” they explained.
A variety of applications
Compared with other tech investment waves, such as SaaS companies and the dot-com era, this AI-driven valuation boom is associated with a more favorable earnings curve for capital. “There is always criticism that much venture investment is based on promises rather than profits. We are not saying the profits are here yet, but the path to real revenues (which eventually leads to earnings) is happening faster than what we have seen in other technology-driven cycles,” they added.
In that vein, Hamilton Lane’s analysis highlights how widespread AI use has become across sectors. Surveying 150 managers across strategies and geographies, they observed an increase in the use of this technology within portfolio companies in the recent past.
In 2024, 44% of surveyed GPs said that between 80% and 100% of their portfolios were actively using AI. The following year, that figure had risen to 61%. Conversely, the percentage of GPs investing in companies with no active use of AI fell from 3% to zero over the same period.
“In one year there has been a significant increase in companies actively using AI. If revenue growth is key for AI to drive expansion, this number is encouraging, although it does not indicate what kind of use is being made of it or at what cost,” the U.S. firm noted in its report.
On the other hand, while they note the share of companies making little use of this technology—3% of GPs reporting that between 0% and 20% of their companies actively use it, and 9% reporting between 20% and 40% usage—this can be explained by the nature of certain sectors.
A plumbing company, for example, will have limited applications for artificial intelligence beyond accounting and logistics. “That lack of direct exposure to AI will be an important consideration for portfolio construction,” Hamilton Lane concluded.
