Crunchbase Data Shows AI’s Most Active Startups Are Becoming Serial Acquirers

A serial acquirer is an AI startup that makes multiple acquisitions instead of treating one purchase as a rare event. M&A becomes a regular growth tool. The company buys products, technology, customer access or specialized teams that fill gaps in its own platform.

The article identifies several repeat buyers. Axiom made 20 AI-related acquisitions across the three-year period, including 10 this year. Legora announced five purchases this year, while Harvey announced four. These deals span areas such as legal research, healthcare data, developer tools and security.

This approach matters because AI companies compete on speed. A startup may acquire a team or product rather than spend months or years building it internally. Serial acquirers can therefore broaden their services quickly, although the article notes that acquisition prices were disclosed for only 12 of 195 deals.

Acquisitions of AI startups by other venture-backed AI companies reached 195 through September 29. That total was 14% higher than the number recorded during all of 2025. The increase came mainly from companies that were already active buyers, rather than from a large expansion in the number of acquiring companies.

The number of buyers grew only 2%. Across the three-year period, 67 repeat buyers generated about 42% of all tracked transactions. Axiom led the field, with 20 AI-related acquisitions overall and 10 during the current year. Paradox and Legora each announced five acquisitions this year.

The pattern shows that dealmaking is becoming concentrated among well-funded, fast-growing startups. These companies are using M&A to expand products and reach customers quickly. The trend may continue if high valuations give startups stock they can use for purchases, but total spending remains difficult to measure because only 12 deal prices were disclosed.

Axiom is the busiest buyer in the article. It made 20 AI-related acquisitions across the three-year period, including 10 this year. This year, Paradox and Legora each announced five acquisitions, Harvey announced four, Sierra and Cursor announced three each, and Delta announced two.

The targets vary widely. Axiom bought companies connected to healthcare data, scientific writing, developer infrastructure, security and specialized talent. Legora and Harvey, both legal AI companies, have pursued businesses adding research, regulatory monitoring and litigation capabilities. Vertical AI startups often buy within their own sectors.

Some purchases were large. Nscale reportedly agreed to acquire Anyscale for $1.65 billion, while Cyera agreed to buy Oasis Security for $1 billion. Paradox bought Coefficient Bio for $400 million. Axiom bought Glass Imaging for $300 million, showing that acquirers seek both products and unusual technical expertise.

An AI startup may buy another company when it needs a capability quickly. Building a product internally requires hiring people, developing technology, testing it and finding customers. An acquisition can combine those assets immediately. The buyer can also enter a market without starting from zero.

Legora CFO David Eckstein described the test directly: does a deal get the company somewhere faster than building it itself? Axiom’s purchases illustrate the range. It acquired Torch Health for health data, Crixet for LaTeX collaboration, Astral for developer tools and Promptfoo for testing AI applications.

The pressure comes from the pace of AI competition. Rama Sekhar of Menlo Ventures said acquiring a team or product is faster than building it. Acquisitions can also bring specialized talent in-house. However, the article gives no evidence that every deal succeeds, and undisclosed prices make the overall financial cost hard to assess.

The acquisition wave is changing how AI companies compete. Instead of building every capability themselves, well-funded startups can buy products and teams that already exist. That may speed product development and help companies reach customers sooner. It could also make repeat buyers more powerful than smaller rivals.

Legal AI shows the mechanism. Acquisitions can combine research, regulatory monitoring and litigation tools into broader platforms. Axiom’s purchases similarly span healthcare data, scientific writing, developer infrastructure, security and specialized talent. Customers may gain access to more services from fewer platforms.

The longer-term result is less certain. More acquisitions could produce faster innovation and wider offerings, but they could also concentrate expertise and products among a small group of buyers. The data already shows concentration: 67 repeat buyers accounted for about 42% of tracked transactions across three years. The article does not report effects on prices or customer choice.

AI startups have several alternatives to buying a company. They can build the missing product with their own engineers, hire individual specialists, form a partnership, license technology or use open-source tools. These approaches may let the startup keep more control and avoid integrating another company’s systems and culture.

The trade-off is speed. Internal development requires time and continuing investment. Hiring people one by one may not bring a complete product, customer base or working team. Partnerships and licenses can provide access without ownership, but they may limit control over the technology, roadmap or commercial relationship. Open-source tools can reduce starting costs but still require expertise to adapt and maintain them.

The article emphasizes why acquisitions are attractive: AI companies want to move quickly, and investors expect rapid growth. M&A can deliver a product and specialized team together. Yet acquisitions also involve negotiation, integration and financial cost, while the article reports prices for only 12 of 195 deals.

Venture funding gives AI startups the capital and financial backing to pursue expansion. High company valuations add another advantage. A highly valued startup can use its shares as currency when buying another company, rather than paying the entire price in cash. This can make a large deal easier to complete.

Rama Sekhar said high valuations have given AI startups cheap currency for acquisitions, with minimal dilution. In this context, dilution means existing investors own a smaller percentage of the combined company after new shares are issued to sellers or other recipients. For example, if more shares are created, each old share represents a smaller slice of ownership.

Stock-funded deals can preserve cash and reduce immediate financing pressure. They may also make acquisitions possible when targets accept shares because they see future value in the buyer. The trade-off is that existing investors’ percentage ownership can fall. The article does not quantify dilution in the reported transactions.