Anthropic is reportedly set to pay Elon Musk’s xAI around $1.25 billion per month (roughly £930 million) for access to computing infrastructure, highlighting the extraordinary costs behind today’s AI race.
The agreement underlines how building advanced artificial intelligence models increasingly depends on access to huge amounts of computing power, often costing more than the models themselves.
Training and running large AI systems requires enormous data centres packed with specialised chips capable of handling complex workloads. As competition between AI firms accelerates, access to these resources is becoming one of the industry’s biggest advantages.
The reported deal would represent one of the largest known spending commitments for AI infrastructure, showing how companies are prioritising scale and performance despite soaring costs.
Anthropic, best known for Claude AI, has grown rapidly in recent years while competing with firms including OpenAI and Google. Meanwhile, xAI has expanded aggressively under Elon Musk, investing heavily in AI chips and data centres.
The arrangement suggests the AI industry is entering a new phase where computing capacity may become as valuable as the models themselves.
Experts increasingly argue that future leaders in artificial intelligence will not only be those with the smartest systems, but also those controlling the infrastructure needed to run them.
The reported spending highlights a striking reality: the AI boom is becoming as much an infrastructure battle as a technology race.
Author: Kieran Seymour