Loading page…
Deal counts are rising much faster than the number of buyers. Companies describe acquisitions as a shortcut to products, talent and customers, but sparse pricing leaves total spending unclear.
Listen to this story
Crunchbase’s Oct. 6 analysis found acquisitions of AI startups by venture-backed AI companies had reached 195 by Sept. 29, 2026—14% above the full-year 2025 count—even though the number of buyers rose only 2%. Repeat acquirers, rather than a broad expansion in the buyer pool, are driving the deal flow as companies add products, technical teams and customer reach. The count measures transactions, not total spending: prices were disclosed for only 12 deals, leaving the reported $1.65 billion Nscale–Anyscale acquisition insufficient to gauge marketwide acquisition costs.
Across the three years Crunchbase reviewed, 67 repeat buyers accounted for about 42% of tracked acquisitions.
OpenAI led with 20 AI-related acquisitions, including 10 in 2026; Anthropic and Legora each announced five.
Legora’s purchases span legal research, regulatory monitoring, commercial real estate analysis and litigation tools.
AI startups are buying their way toward faster growth, but the surge reflects more deals by recurring buyers—not a comparable surge in new acquirers. Crunchbase’s October 6 analysis counted 195 acquisitions of AI startups by other venture-backed AI companies through September 29, 2026, already 14% above the total for all of 2025.
The number of buyers grew just 2%, pointing to repeat acquirers as a major driver of the increase. Across the three-year period Crunchbase reviewed, 67 repeat buyers accounted for about 42% of tracked transactions. OpenAI led with 20 AI-related acquisitions, including 10 this year.
Crunchbase counted acquisitions through September 29, exceeding the full-year 2025 total before the final quarter began.
Only 12 transactions carried disclosed prices, limiting what the dataset reveals about total acquisition spending.
The purchases show what buyers want to add. Legora’s five acquisitions span legal research, regulatory monitoring, commercial real estate analysis and litigation tools. One target, Walter AI, builds agents that work inside Microsoft Word and Outlook. Another, Wexler AI, helps litigation teams extract facts and build case timelines.
Harvey COO Katie Burke told Crunchbase News that the company seeks technical talent with deep legal or adjacent industry experience. Its purchase of Benchmark, which helps asset managers reuse insights from previous investments, extended Harvey’s presence in asset management. Burke described the acquisition strategy as selective but aggressive.
OpenAI’s purchases ranged from healthcare data and scientific-writing software to developer infrastructure and security tools. Ona, formerly Gitpod, supplies cloud environments where developers and AI agents can keep working after a laptop closes. Ona’s own announcement confirms its transaction closed on August 10, joining OpenAI’s Codex team.
Targets also describe selling as a way to accelerate, rather than a retreat from weak demand. TakeOff founder Aakash Thumaty wrote that his three-person company had capital, customers and a near eight-figure revenue run rate. He said joining Sierra offered a chance to pursue its vision at greater scale.
Crunchbase identified Nscale’s reported $1.65 billion acquisition of Anyscale as the largest transaction with a recorded price. But with prices disclosed for only 12 deals, the analysis cannot provide a reliable picture of overall spending. Its dataset also excludes transactions that have not been publicly reported.
Loading discussion...
Join the conversation
Explain when greater scale is worth giving up independence.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.