Amazon bets another $4B on Anthropic

Amid concerns about the return of AI winter, when funding and advancements slow down, neural-network golden child Anthropic reports the doubling of Amazon's already substantial financial backing.

It was only back in March that Amazon finished shoveling $4 billion into Anthropic, maker of the Claude family of generative models. Now comes word of another $4 billion, bringing the e-commerce and cloud giant's total investment to – check the math – $8 billion.

In the face of so much cash – no cloud credits were involved, Amazon told us previously – Anthropic has declared Amazon Web Services to be its "primary cloud and training partner." That represents an expansion of the previously declared relationship in which AWS was described only as the "primary cloud partner."

Presumably Google, which committed to investing a mere $2 billion in Anthropic last year, must accept some lesser form of relationship.

Tie-ups like these piqued the interest of the US Federal Trade Commission in January, based on concerns that deals between tech giants and AI firms may limit competition. The watchdog agency hasn't announced any action since then beyond issuing a joint statement in July with EU and UK regulators about the need to promote competition among AI firms.

So far, nothing has come of the regulatory concern. In August, the UK Competition and Markets Authority (CMA) opened an investigation into Amazon's arrangement with Anthropic. A month later, the competition watchdog dropped its investigation. In October, the CMA began sniffing around Google's investment in Anthropic.

• Whomp-whomp: AI PCs make users less productive
• Kyndryl insiders say there's little new business
• Prepare for an AI policy upending under Trump, say experts
• Datacenters could blow up your electric bill thanks to AI

Anthropic says it expects to work with Annapurna Labs at AWS to develop and refine AWS' Trainium accelerators, which help speed up machine learning workloads.

"Through deep technical collaboration, we’re writing low-level kernels that allow us to directly interface with the Trainium silicon, and contributing to the AWS Neuron software stack to strengthen Trainium," Anthropic said. "Our engineers work closely with Annapurna’s chip design team to extract maximum computational efficiency from the hardware, which we plan to leverage to train our most advanced foundation models."

Part of the rationale for the deal is that business customers – as opposed to more reticent consumers – appear to be enthusiastic about AI services. Anthropic notes that Pfizer is using Claude models in the model management service Amazon Bedrock to accelerate medical research and reduce operational costs. Other customers include Intuit, Perplexity, and the European Parliament, which is said to use Claude in its Archibot document search and analysis service.

"The response from AWS customers who are developing generative AI applications powered by Anthropic in Amazon Bedrock has been remarkable," AWS CEO Matt Garman said in a statement. "By continuing to deploy Anthropic models in Amazon Bedrock and collaborating with Anthropic on the development of our custom Trainium chips, we’ll keep pushing the boundaries of what customers can achieve with generative AI technologies." ®