The news: Google is building various AI products hinged on its DeepMind and Gemini models, but it has also invested in AI startups, including Anthropic, which Google now has a 14% stake in, The New York Times revealed.
Google is an Anthropic hosting partner and investor but has no voting rights or board seats. With over $3 billion already invested and another $750 million set for September, Google is expanding its stake and influence in one of AI’s fastest-growing startups.
Why Anthropic? Google has many AI startups to choose from—Perplexity, a potential search rival, aligns more closely with its technology, while OpenAI remains firmly tied to Microsoft. But Anthropic has been on Google’s radar since 2023, when it became a Google Cloud partner.
Google likely chose Anthropic due to its AI chatbot Claude’s superior programming skills—last month, Claude 3.7 Sonnet scored 70.3% on a benchmark testing AI’s real-world software engineering and problem solving abilities. That was far ahead of OpenAI’s o1 (48.9%) and DeepSeek-R1 (49.2%). It also excels in agentic tasks, per Venture Beat.
Google sees AI code creation as a key differentiator for itself and its users.
Less than 10% of enterprise software engineers used generative AI (genAI) code assistants in early 2023—that number is expected to grow to 75% by 2028, according to a Gartner poll.
Mass adoption of Gemini Code Assist could help Google secure the industry standard and build an audience for future monetization.
Why not buy the whole thing? Google’s ongoing antitrust cases will make any high-profile AI acquisition difficult, if not impossible, especially since it’s considered one of the top AI players. Its cautious collaboration and ongoing cloud partnership with Anthropic keep it strategically positioned in AI’s evolution.
Our take: By maintaining low-stakes investments, Google can selectively back projects with the highest potential impact. It’s a smart strategy that could be emulated by other Big Tech companies falling behind in AI development.