Why Is Google Taking the Core of AI Companies Instead of Buying Them Whole?

Among the companies listed in this year’s EU Digital Markets Act gatekeeper acquisition disclosures, Google stands out.

Google’s activity is difficult to read as a simple list of mergers and acquisitions. Looking only at the 2026 entries, Google is absorbing AI agents, cloud security, and 3D generative AI in different ways. Some transactions are traditional acquisitions. Others involve hiring specific engineers and licensing intellectual property on a non-exclusive basis.

This shows how Big Tech expansion is changing in the AI era.

In the past, the key question was which company had been acquired. But in the AI industry, there may be things more important than the company as a whole: key researchers, engineering teams, model implementation experience, data-integration methods, agent-orchestration know-how, and experience applying AI systems to enterprise customers.

This year, Google appears to be targeting precisely those core pieces.

The most symbolic case is Contextual AI.

According to the EU DMA list, Google notified a transaction involving the hiring of certain software engineers from Contextual AI and a License and Release Agreement that included a non-exclusive license to relevant Contextual AI intellectual property. Contextual AI is described as an enterprise software company that helps businesses create and deploy AI agents based on their own data through retrieval-augmented generation, or RAG.

This is the important point.

What Google is trying to obtain is not merely chatbot technology. It is the capability to build and deploy AI agents grounded in enterprise data. In the early phase of generative AI, the competition centered on general search, writing assistance, and code generation. Now the competition is moving toward agents connected to enterprise data.

Enterprise customers do not simply want general-purpose AI. They want AI connected to their own documents, knowledge bases, work systems, policies, and customer information.

Contextual AI sits directly at that point.

From the perspective of Google Cloud, this is strategically natural. When enterprises adopt AI, the biggest problem is often not the model itself. It is data access and business application. Internal data is fragmented. Permission systems are complex. Answers need sources and context. A RAG-based agent platform is a core technology for solving this problem.

But Google does not describe this as a full acquisition of Contextual AI.

The notification states that the transaction is based on Alphabet’s view that it does not constitute a concentration under the EU Merger Regulation because it does not involve the acquisition of a business to which turnover can be attributed.

This is the core of Google’s movement this year.

In the AI era, a company does not have to buy another company in full to obtain core competitive capabilities. It can hire the people it needs, access the IP it needs, and combine the necessary technology with its own cloud and products. Traditional M&A regulation focuses on equity and control, but the substance of AI competition may lie in the movement of people and technology.

Google’s second axis is cloud security.

In February 2026, Google notified its acquisition of Wiz. Wiz is a New York-based cloud security company active in the cloud-native application protection platform, or CNAPP, market.

This transaction is different from Contextual AI.

Contextual AI represents AI-agent application capability. Wiz represents cloud-security capability. But the two are not separate. As enterprise AI spreads, cloud security becomes more important. AI agents access enterprise data, call applications, and operate inside business systems. In that context, cloud vulnerabilities, permission misconfigurations, data exposure, and runtime risks become AI-adoption risks.

Google Cloud competes against AWS and Microsoft Azure in the enterprise cloud market.

In that competition, AI models alone are not enough. Enterprises want to use AI, but they also worry about security and compliance. Cloud-security capabilities such as Wiz can strengthen the trust foundation Google Cloud needs in order to attract AI workloads.

In other words, Google’s 2026 strategy is not simply “make better AI.”

It is closer to “make enterprise AI safer to use.”

For AI agents to access enterprise data, RAG and agent technologies are needed. For those agents to operate safely in the cloud, cloud security is needed. Contextual AI and Wiz show these two axes.

The third axis is 3D generative AI.

In January 2026, Google notified the acquisition of Common Sense Machines, or CSM. CSM is described as a technology startup that uses generative AI models to create three-dimensional digital objects, enabling 3D content generation for design, gaming, and visualization applications.

This entry shows that Google’s AI competition is not limited to text and image generation.

3D generative AI can connect to gaming, advertising, design, architecture, robotics, XR, spatial computing, and digital twins. If 3D objects can be created from text prompts and placed into simulations or interactive content, the cost and speed of content creation can change significantly.

Google already has Search, Maps, Android, YouTube, Cloud, Gemini, and Workspace. If 3D generation technology is combined with these assets, its potential use cases become wide. 3D asset generation is not merely a creative tool. It may also become part of the spatial internet and the ability of AI agents to understand environments.

When Google’s three transactions are placed side by side, the direction becomes visible.

Contextual AI is about AI agents based on enterprise data.

Wiz is about cloud security.

Common Sense Machines is about 3D generative AI.

These three may look like different markets, but they point toward one larger direction.

Google is trying to become not merely an AI model company, but an AI execution-infrastructure company. It wants to build models, connect them to enterprise data, operate them safely in the cloud, and expand generative interfaces from text and images into 3D worlds.

The Contextual AI transaction is especially important.

It shows a new mode of Big Tech absorption in the AI era. Google did not say it was acquiring all of Contextual AI. It hired some engineers and secured an IP license. It also maintained the position that the transaction was not a concentration under the EU Merger Regulation.

This is a regulatory-sensitive structure.

On the surface, it is talent mobility.

In substance, it may be the movement of core technical capability.

On the surface, it is a non-exclusive license.

But once Big Tech combines that license with its own cloud and AI products, the market impact can become significant.

On the surface, it is not the acquisition of a business.

But the independent startup’s core growth engine may be weakened.

This is why the EU DMA puts such transactions into a public list.

AI talent and IP absorption deals are increasingly difficult to capture through traditional merger-notification thresholds. Even if a startup has little current revenue, it may hold technology and talent that matter for future competition. By requiring gatekeepers to notify such digital-sector transactions, the DMA enables regulators and markets to see the pattern.

Google’s activity this year is therefore not a simple shopping list.

It is a signal of where Google sees the bottlenecks in the AI market. Model-performance competition is already intense. But to sell AI to enterprise customers in practice, three additional layers are needed.

First, agents connected to enterprise data.

Second, cloud security that allows those agents to operate safely.

Third, multimodal extensibility beyond text and images into 3D worlds.

Google is securing all three pieces this year.

That does not mean success is guaranteed.

Contextual AI-style talent and IP absorption may continue to draw regulatory attention. The Wiz acquisition may raise questions about competition and integration risks in the cloud-security market. The CSM acquisition will have to prove whether 3D generative AI can be translated into real products and revenue.

But the direction is clear.

Google does not see AI as a single model.

It sees AI as a full stack connecting enterprise data, cloud security, developer tools, generative content, and spatial computing.

That strategy is visible in this year’s DMA list.

What matters most is that Google is not handling everything through large acquisitions. Some companies it buys outright. From others, it hires people. From others, it licenses IP. It is a combinational strategy: take the technology in the form that best fits the need.

Big Tech acquisition strategy in the AI era is becoming more subtle and more precise.

Even if a company name does not disappear, key talent can move.

Even if a product remains formally independent, its IP can enter Big Tech products.

Even if a transaction is said not to be a concentration, the structure of market competition can change.

That is why Google’s transactions this year matter.

Contextual AI, Wiz, and Common Sense Machines may each look like separate points. But when the three points are connected, they reveal the AI map Google is drawing: build enterprise AI agents, operate them securely in the cloud, and extend generative AI into the 3D world.

Google is not merely buying AI companies.

It is quietly absorbing the future work environment, cloud infrastructure, and content-generation space in which AI will operate.