one year on
Amazon to invest up to $4 billion in AI startup Anthropic
The e-commerce giant takes a minority stake in the OpenAI rival, naming AWS as its primary cloud provider and deepening the Silicon Valley trend of hyperscaler patronage for frontier labs.
Amazon announced Monday it will invest up to $4 billion in Anthropic, taking a minority stake in the two-year-old AI startup founded by former OpenAI executives. The deal also makes Amazon Web Services Anthropic’s primary cloud provider for mission-critical workloads, including model training and deployment on AWS’s custom Trainium and Inferentia chips.
The agreement is part of a broader collaboration to develop reliable foundation models. Anthropic will expand its support for Amazon Bedrock, giving AWS customers early access to fine-tuning capabilities for its Claude 2 chatbot. The startup’s 100,000-token context window and safety research are key selling points for enterprise clients like LexisNexis, Bridgewater Associates, and Lonely Planet.
The tie-up solidifies a growing industry pattern: Microsoft’s bet on OpenAI, Google among Anthropic’s backers, and now Amazon’s direct investment. The deal extends a pattern of big cloud providers backing frontier AI labs. The deal also signals Amazon’s push to show that there are alternative chips that can be used for AI applications by promoting its own silicon. Anthropic says its corporate governance remains unchanged under its Long Term Benefit Trust.
The record
One year later — open only if you can handle spoilers
The investment helped cement Anthropic's position as a top AI contender, with Claude models becoming a core offering on AWS. Amazon's chip bet paid off slowly, as Trainium adoption grew but Nvidia retained the lion's share of AI workloads through 2025.