Some of the AI industry’s fastest-growing startups are becoming serial acquirers, buying smaller companies to fill product gaps, enter new markets and bring specialized teams in-house, a review of Crunchbase data shows. While AI giant OpenAI is by far the busiest of these buyers, well-funded startups in legal tech, customer service and software development have also made multiple acquisitions this year.
The buying spree has pushed acquisitions of AI startups by other venture-backed AI companies to 195 through Sept. 29, according to Crunchbase data — 14% more than in all of 2025. Yet the number of buyers grew just 2%, indicating that increasingly active acquirers are driving much of the increase.
Those deals point to a new phase of competition in the AI sector: Well-funded startups are using M&A to broaden their products and reach new customers faster than they could by building everything themselves. In legal tech, for example, acquisitions are bringing research, regulatory monitoring and litigation tools into broader platforms.
Legora CFO David Eckstein described that calculus in a LinkedIn post earlier this year explaining the company’s back-to-back purchases of multiple startups: “M&A is explicitly part of how we accelerate what we’re building. The question we always ask is: does this deal get us somewhere faster than we’d get there ourselves?”
Repeat buyers step up dealmaking
Some of the AI acquirers have returned to the dealmaking table more than once in recent years, Crunchbase data shows. Across the three-year period, 67 repeat buyers accounted for about 42% of all transactions tracked. OpenAI was by far the most active, with 20 AI-related acquisitions, including 10 this year.
This year’s other repeat buyers include Anthropic and legal AI startup Legora, which have each announced five acquisitions, followed by legal AI startup Harvey with four. Customer-service AI provider Sierra and coding company Cursor have each made three acquisitions this year, while Cohere announced two. Many of the buyers are vertical AI startups acquiring companies in their respective areas.
Several of this year’s transactions are sizable. Nscale’s reported $1.65 billion acquisition of Anyscale was the largest with a recorded price, followed by Cyera’s $1 billion acquisition of identity security startup Oasis Security.
Anthropic’s acquisition of Coefficient Bio, which develops AI for pharmaceutical research, was valued at $400 million. OpenAI’s $300 million acquisition of Glass Imaging, a Los Altos, California, startup that uses customized AI in camera hardware for computational photography, and Sword Health’s acquisition of Kaia Health, valued at up to $285 million, rounded out the five largest deals recorded in the dataset through Sept. 29.
Overall, prices were disclosed for only 12 of the 195 deals, making it difficult to gauge how much AI startups are spending on acquisitions overall.
Dealmaking has clearly picked up as more AI companies turn to acquisitions — and, in some cases, make them a recurring part of their growth strategy. 1
“It’s all about speed in the AI world,” according to Rama Sekhar, partner at Menlo Ventures, which has backed numerous AI startups including Anthropic and Legora. “It’s faster to acquire a team or product than build it yourself. If you’re not growing 10x, you’re not interesting to growth investors, which leaves a gap in the funding market for AI startups that need a home. High valuations have also given AI startups cheap currency to use their stock to get these deals done with minimal dilution.”
What AI startups are buying — and why
The specific companies being acquired offer insight into what AI startups hope to gain from their stepped-up M&A activity.
OpenAI’s purchases have ranged widely, from healthcare data and scientific-writing software to developer infrastructure, security tools and specialized talent. The San Francisco-based company announced three acquisitions in January alone, setting the tone for a busy M&A year.
In January, it acquired:
- Convogo, which built AI software to help executive coaches automate leadership assessment reports.
- Torch Health, an AI-powered health app that aims to unify scattered medical records from hospitals, labs, wearables and consumers.
- Crixet, which provides LaTeX editing, error detection and team collaboration.
In February, OpenAI participated in an acqui-hire deal involving open-source AI agent OpenClaw and its creator, Peter Steinberger. Then in March, it announced plans to acquire Astral, a creator of open-source tools for software developers. It also snapped up Promptfoo, an open-source tool for testing AI applications.
In June, OpenAI agreed to buy Ona, formerly known as Gitpod, which provides secure cloud environments where developers — and, increasingly, AI agents — can continue working after a user’s laptop is closed. In August, it also picked up Instant, an AI presentation company that converts prompts, notes and documents into editable presentations.
Vertical AI roll-ups
Well-funded vertical AI startups, particularly in legal tech, have also been busy this year buying smaller companies, an analysis of Crunchbase data shows.
San Francisco-based Harvey’s AI-related purchases have focused on filling gaps around its core platform. Hexus developed tools for creating product demos, videos and guides, while Lume built software that helps companies connect customer data and applications with AI systems.
New York-based Benchmark developed software that helps asset managers capture insights from previous investments and apply them to new deals. That acquisition helped expand Harvey’s presence in the asset management space. Its fourth known acquisition this year, announced Sept. 9, was Guardrails AI, which builds open-source tools for testing, monitoring and managing AI agents.
Katie Burke, COO of Harvey, told Crunchbase News that the company’s M&A strategy is rooted in finding “technical talent with high ownership and deep experience in legal tech or an adjacent space to legal.”
She pointed to Benchmark as an example. The co-founders, she said, “know the asset management space cold, and their name was dropped so many times in customer conversations that it was a natural fit for them to join our team.”
Burke further described the startup’s M&A strategy as “selective but aggressive.”
“We hold an incredibly high bar for talent and when we identify an additive company, we move quickly and will continue to do so this year and beyond,” she added.
Legora has pursued an even broader legal-tech rollup. The Stockholm-based company’s five announced acquisitions this year include:
- Walter AI, a nine-person Canadian startup whose agents work inside Microsoft Word and Outlook;
- Qura, a Stockholm company building AI-native legal research tools;
- Graceview, an Australian regulatory-intelligence platform that monitors more than 100 areas of law for changes;
- Cadastral, whose AI agents analyze documents and data for commercial real estate companies; and
- London-based Wexler AI, which helps litigation teams extract facts from large collections of documents, identify inconsistencies and build case timelines.
Sierra’s acquisitions reflect an effort to expand both geographically and beyond conventional customer-service automation. In March, it acquired Tokyo-based enterprise AI startup Opera Tech as part of its expansion in Japan, followed by Paris-based Fragment, which helps businesses automate operational work with AI. In July, Sierra bought TakeOff, a 14-month-old startup developing “long-horizon” agents.
For acquisition targets, joining a larger platform can offer a shortcut to scale even when their own businesses are growing rapidly. TakeOff founder Aakash Thumaty wrote in a July blog post that his company was posting a “near 8-figure run rate” with a team of just three people. “The advice for an AI startup growing at our pace is to hire out a sales team, raise again, and keep going,” he wrote. “We had capital, customers, and great traction,” but the Sierra acquisition offered an opportunity to “accelerate our shared vision and simultaneously build it at a grander scale.”
The surge in M&A activity is driven by both a wider pool of buyers and increasingly active serial acquirers. At this pace, we should expect even more in the months and years ahead.
Related Crunchbase queries:
Related reading:
- Data: OpenAI Has Already Done Nearly As Many M&A Deals In 2026 As It Did All of Last Year
- Startups Are Still Acquiring Startups, Led By Ultra-High-Valuation Unicorns
- Your AI Strategy May Be Destroying Your Exit Value
Illustration: Dom Guzman
Our analysis is based on transactions in the Crunchbase dataset and doesn’t include deals that haven’t been publicly reported↩
