This year, established tech giants have dominated the much-hyped generative AI space by investing far more in generative AI startups than venture capital groups.

Recent data from private market researchers PitchBook shows that two-thirds of the $27 billion raised by early-stage AI companies in 2023 came from a series of blockbuster deals struck by Microsoft, Google, and Amazon last year.

The massive expenditure, which skyrocketed following the November 2022 release of OpenAI’s ChatGPT, demonstrates how Silicon Valley’s largest organizations are displacing more established tech investors in search of the largest transactions in the sector.

Leading Silicon Valley investors have also shown interest in the emergence of generative AI—systems that can create human-like text, images, audio, and video in a matter of seconds. However, VCs are now outmatched because they had to reduce their expenditures in response to rising interest rates and declining valuations for the companies in their portfolio.

“Over the past year, we’ve seen the market quickly consolidate around a handful of foundation models, with large tech players coming in and pouring billions of dollars into companies like OpenAI, Cohere, Anthropic and Mistral,” said Nina Achadjian, a partner at US venture firm Index Ventures referring to some of the top AI startups.

“For traditional VCs, you had to be in early and you had to have conviction—which meant being in the know on the latest AI research and knowing which teams were spinning out of Google DeepMind, Meta and others,” she added.

A series of agreements, including billions of dollars raised by San Francisco-based Anthropic from both Google and Amazon, and Microsoft’s $10 billion investment in OpenAI, helped push overall spending on AI groups to nearly three times the previous record of $11 billion set two years ago.

In 2021, venture capital investment in the technology sector reached all-time highs as investors made massive investments across a variety of industries, especially those most affected by Covid-19, by taking advantage of record-low interest rates.

It takes a great deal of time, money, and processing power to develop and train generative AI tools. Because Big Tech companies can offer cloud infrastructure, access to the most powerful chips, and financial resources, start-ups have chosen to collaborate with them.

This has caused the valuations of private start-ups in the industry to soar, making it more difficult for venture capitalists to back the cutting edge of technology. OpenAI is attempting to almost triple its earlier-this-year valuation of $86 billion through the sale of employee stock.

“Even the world’s top venture investors, with tens of billions under management, can’t compete to keep these AI companies independent and create new challengers that unseat the Big Tech incumbents,” said Patrick Murphy, founding partner at Tapestry VC, an early-stage venture capital firm.

“In this AI platform shift, most of the potentially one-in-a-million companies to appear so far have been captured by the Big Tech incumbents already.”

But VCs are still present in the market. Josh Kushner’s New York-based company, Thrive Capital, is the principal investor in OpenAI’s employee stock offering. The company was previously backed by Thrive earlier this year. Thrive made investments in 2023 even though venture capital spending declined.

Since its founding in May of this year, Mistral, a startup based in Paris, has raised about $500 million from investors, including chipmaker Nvidia and venture firms Andreessen Horowitz and General Catalyst.

Certain venture capitalists are aiming to fund businesses that are creating applications on top of the “foundation models” created by OpenAI and Anthropic. This approach is similar to how applications were first created for mobile devices in the years following the release of smartphones.

“There is this myth that only the foundation model companies matter,” said Sarah Guo, founder of AI-focused venture firm Conviction. “There is a huge space of still-unexplored application domains for AI, and a lot of the most valuable AI companies will be fundamentally new.”