Alphabet Inc.’s Google is in early discussions to supply Anthropic PBC with cloud computing capacity valued in the high tens of billions of dollars, a potential expansion that would deepen one of the most consequential partnerships in the artificial intelligence industry, Bloomberg reported.
The agreement under discussion would allow Anthropic to access Google’s tensor processing units (TPUs), the company’s custom-built chips designed to accelerate large-scale machine learning operations. People familiar with the matter said the talks remain preliminary and may not lead to a final deal.
Founded in 2021 by former OpenAI researchers, Anthropic has positioned itself as one of a key players in the development of “safe and reliable” AI systems. Its flagship language model, Claude, competes with OpenAI’s ChatGPT and is already available through Google Cloud’s Vertex AI platform.
Google became both an investor and cloud provider for Anthropic following an expanded partnership announced in late 2023. The collaboration gave Anthropic access to Google Cloud’s high-performance computing infrastructure, including the TPU v5e accelerators. Anthropic said at the time that Google’s infrastructure would help it “deploy AI systems to a larger set of people.”
The company also maintains a parallel relationship with Amazon Web Services, which invested up to 4 billion US dollars in Anthropic and serves as its primary cloud provider for mission-critical workloads. This dual-cloud arrangement prompted the UK Competition and Markets Authority to open an inquiry in mid-2024 over potential competition concerns.
Neither Google nor Anthropic has commented publicly on the current negotiations. In after-hours trading, Alphabet shares rose more than 3.5 percent, while Amazon.com Inc., another Anthropic backer, fell about 2 percent.
If finalised, the deal would be among the largest cloud computing commitments in the AI sector, reflecting the scale of infrastructure investment now required to train and deploy next-generation generative models.
