In October 2023, Google agreed to invest up to $2 billion in Anthropic, the San Francisco-based AI safety company founded in 2021 by former OpenAI researchers. The commitment consisted of an initial $500 million cash investment, with an additional $1.5 billion structured as convertible notes tied to future milestones. The transaction accompanied a multi-year cloud computing agreement under which Anthropic would use Google Cloud infrastructure and Google's custom Tensor Processing Units (TPUs) to train and serve its Claude (AI model family) family of large language models.
The investment was reported publicly in late October 2023, shortly after Amazon's agreement to invest up to $4 billion in the company. Together, the two deals made Anthropic one of the most heavily capitalized AI startups of the period, and illustrated the willingness of major cloud providers to fund model developers in exchange for long-term compute commitments.
Background
Anthropic was founded in January 2021 by seven former OpenAI employees, including siblings Daniela Amodei and Dario Amodei, who serve as president and CEO respectively. The company positioned itself around AI safety research, including work on Mechanistic interpretability, alignment, and the societal impacts of Artificial intelligence. It raised $124 million in May 2021 and released its first publicly available model, Claude 1, in March 2023.
Google had been an early investor in Anthropic. The company participated in Anthropic's 2022 Series B round, and Google Cloud was named as Anthropic's preferred cloud provider at that time. Google's interest in Anthropic sat alongside its ownership of DeepMind, acquired in 2014, and its own Bard and later Gemini model programs. The October 2023 investment therefore represented both a financial bet and a commercial arrangement intended to secure a fast-growing AI workload for Google's cloud business.
Terms of the deal
The reported structure of the investment was unusual for a corporate venture deal. Google committed $500 million upfront, with the remaining $1.5 billion to be delivered through convertible notes that would convert into equity upon specified conditions. This milestone-based structure resembled the arrangement Google had used in earlier Anthropic rounds and mirrored the structure of Amazon's concurrent investment.
In exchange, Anthropic agreed to use Google Cloud as a primary provider for training and inference. The company gained access to Google's TPU accelerators, an alternative to the NVIDIA GPUs that dominated AI training at the time. Anthropic had previously trained models on a mix of hardware, and the TPU agreement gave it a second large-scale compute source alongside its existing arrangements.
Neither company disclosed the precise valuation at which the convertible notes would convert. As of the announcement, Anthropic's valuation was reported in the range of $20 billion to $30 billion, following earlier 2023 funding activity.
Strategic context
The deal formed part of a broader pattern in 2023 and 2024 in which large cloud and chip companies took significant stakes in frontier model developers. Microsoft (AI) had invested heavily in OpenAI, and Amazon Web Services committed up to $4 billion to Anthropic in September 2023. Google's investment made it Anthropic's second major strategic backer among the hyperscalers.
For Google, the arrangement served several purposes. It tied a leading model developer to Google Cloud, providing a large and growing customer for TPUs. It also gave Google visibility into Anthropic's technical direction, though the companies maintained separate model development efforts. Google's own AI strategy at the time was in flux: the company had launched Bard in March 2023 and was preparing the Gemini model family, announced in December 2023.
For Anthropic, the investment supplied capital for Compute-intensive training runs and reduced reliance on a single cloud provider. The company had emphasized that its safety mission required it to remain at the frontier of model capability, which in turn required access to large-scale compute.
Relationship to Anthropic's funding history
The Google investment was one of several large rounds Anthropic raised between 2023 and 2025. It followed the 2023 funding round and the Amazon investment, and preceded the 2024 funding round, the 2024 Safari round, and later rounds including the 2025 $3.5 billion round. Google participated in multiple subsequent Anthropic financings, and its stake grew as the company's valuation rose.
In October 2025, Anthropic announced an expanded cloud partnership with Google that gave it access to up to one million Google TPUs, potentially adding more than one gigawatt of AI compute capacity by 2026. That agreement built on the relationship established by the 2023 investment.
Reception and analysis
The investment drew attention because it placed Google in the position of funding a company founded by former OpenAI employees while simultaneously competing with it through DeepMind and Gemini. Commentators noted the potential for conflicts of interest, particularly around access to information about a competitor's research, though both companies described the relationship as a standard investor-portfolio arrangement.
The deal also reinforced concerns about concentration in the AI industry, in which a small number of cloud providers and chip suppliers held stakes in the leading model developers. Analysts observed that the convertible-note structure allowed Google to limit its initial outlay while retaining the option to increase its stake if Anthropic met performance or commercial milestones.
Aftermath
Anthropic continued to release models through 2023 and 2024, including Claude 2 and the Claude 3 family. The company expanded its enterprise offerings and developer tools, later introducing the Model Context Protocol and Claude Code. Its researcher roster grew to include figures such as Chris Olah, Jan Leike, Amanda Askell, and Andrej Karpathy.
Google's investment remained part of a wider set of strategic relationships for Anthropic, which by 2025 also included Microsoft and Nvidia commitments reported at up to $15 billion, alongside a $30 billion Azure compute purchase. The 2023 Google deal is generally treated as the point at which Anthropic's cloud partnerships became a central element of its corporate strategy.