Summary:
Many companies feel pressure to engage with AI. One visible way to do this is to buy firms that work with AI. Our Journal of Management Studies article, “Examining the effect of a firm’s AI specialization on the technology firms it acquires: A real options perspective,” looks at who tends to acquire AI targets and when those acquisitions are more likely to occur. Using U.S. public firms from 2004 to 2015, we find that firms with a higher share of AI patents in their own portfolio are more likely to acquire AI targets. We also see three target attributes linked to a higher likelihood of an AI deal: 1) lower R&D intensity, 2) clear evidence of turning ideas into products, and 3) greater product-market overlap with the buyer.
Our Motivations:
Whenever leaders are looking for opportunities for their firms, one prominent option is to make an acquisition. This is especially true when faced with new technologies and uncertainties. Consequently, when considering how to help their firms deal with the realities of AI, Leaders often ask, “Should we buy an AI company?” The appeal is speed, since a deal can be faster than building capability from scratch. But acquisitions entail a high degree of uncertainty, and it is often challenging to integrate the acquiring and the target firms. These challenges are likely to be heightened with AI acquisitions due to the emergent and quickly evolving nature of the technology. We set out to learn how firms navigate the tension of wanting to enhance AI capabilities while also dealing with the heightened challenges when buying AI capabilities, focusing on which acquirers are likely to pursue AI acquisitions and what type of targets they pursue.
What We Looked At:
Two questions guided our study:
- Who buys? Are companies that have developed more AI technologies themselves more likely to acquire AI targets?
- Who they tended to acquire? We examined whether experience with AI technologies impact the characteristics of the firms they acquire. We focused on features that could reduce some of the uncertainty of undertaking an AI acquisition, including the intensity of research and development, signs of regular productization (moving from ideas to workable offerings over time), and the closeness of customers or products between buyer and target.
What We Found about AI Acquisition:
- Buyers with more AI know-how are more likely to acquire AI targets.
Companies with a larger share of AI-related inventions in their own portfolio are more active in acquiring AI targets. This finding is consistent with the argument that having AI knowledge reduces the perceived uncertainty in acquiring an AI firm.
- Three target features are linked with a higher chance of an AI deal.
When AI-specialized buyers acquire, they look to acquire AI assets they can internalize and exploit quickly. Specifically, we find they tend to target firms with the following attributes:
- Lower research and development intensity. Targets with lighter research spending (relative to sales) appear more often in the AI deals we observe. This profile often reflects firms that are further along in moving from exploration to offerings.
- A visible habit of turning ideas into products. Targets that regularly convert know-how into usable features through ongoing product development show up more in these transactions.
- Overlap in customers or products. Buyer and target are closer in what they sell or to whom they sell. This proximity is a frequent feature in AI acquisitions and suggests that channel, customer, and product language already align.
Viewed together, internal AI specialization on the buyer side goes hand-in-hand with a higher likelihood of acquiring AI targets, and deals tend to feature targets that have more mature technologies, look more product-ready, and are closer to the buyer’s markets. For example, think of a large cloud provider with many AI teams in-house acquiring a smaller firm that sells ready-made AI modules for fraud detection to the same banks and insurers the provider already serves. The target spends a smaller share of its budget on basic research and more on maintaining and extending products that are already in customers’ hands, so the acquirer can plug these tools into its existing platform and customer channels pretty quickly.
In contrast, though firms with limited experience with AI are less likely to acquire AI firms, when they do, they appear to look to targets that allow them to broadly explore the AI space. This is reflected in targets that are higher in R&D intensity, less likely to have turned technology into products, and have relatively low levels of overlap in customers or products. Imagine a traditional manufacturing firm with little AI work acquiring a small AI lab that has many researchers, several pilots, and a few stable products, and that operates in a market the acquirer has not yet entered. The goal of such a deal is less about immediate integration and more about learning where AI could matter for the business from a more exploratory perspective.
Implications For Different Stakeholders
- Managers
If your firm already does more AI-related work, you are in the group more likely to acquire AI targets. In the deals we observe, targets often have lighter research spending, visible product release activity, and overlap with the buyer’s customers or products. These are useful signals of when deals tend to occur and which profiles commonly pair together. They can guide planning cycles, pipeline watchlists, and internal alignment on where to focus attention. If your firm has little AI experience, the targets you consider may have higher R&D, fewer finished products, and less overlap, allowing the inexperienced firm to explore the AI space. It is crucial to have a fair risk assessment in the acquisition process.
- Boards and investors
A higher share of AI work inside a buyer is linked with greater acquisition activity in this area. Tracking buyer specialization along with the three target signals can support timing discussions (when activity may rise), resource allocation (where to build optionality), and agenda-setting (which kinds of targets are more likely to come forward). This framing helps structure oversight without requiring deep technical dives. If a buyer has little AI experience, any AI deal is more exploratory. Make sure to approve only if the team shows ready data, a steady release plan, and a real customer fit to reduce the uncertainty involved.
Target companies
- Firms that show steady productization, maintain more modest research intensity, and sit closer to a buyer’s customers or products appear more often in AI acquisitions. Understanding how your profile maps to these dimensions can clarify which potential buyers may be more receptive and when outreach or partnership talks are most likely to gain traction. It can also shape how you present your progress to the market. If the acquirer has fewer AI inventions, expect broader exploration. Share your roadmap and recent releases. Lower risk with a pilot, a step-by-step product plan, and a target date for the first usable feature. Match your pitch to the buyer’s readiness.
