Anthropic Reportedly in Talks for Second $45 Billion Compute Deal With Nscale

Anthropic is reportedly negotiating a second major compute deal with cloud provider Nscale worth as much as $45 billion, according to TechCrunch and Bloomberg—though neither company has confirmed the deal as final.

Key points:

• This would be in addition to Anthropic's earlier compute arrangement with Nscale, dramatically expanding the computing capacity Anthropic can draw on to train and run its Claude models.

• The reporting describes the deal as still "in talks" rather than signed, so terms could change or the deal could fall through.

A deal this size reflects how capital-intensive frontier AI has become—compute access, not just algorithmic breakthroughs, is now one of the clearest competitive moats between labs.

The timing is notable: this news arrived in the same week OpenAI disclosed it can't explain its own agents' behavior and researchers exposed the Cairn hacking tool. Investors appear to be pricing in AI's long-term trajectory largely independent of this week's safety headlines, which signals that market confidence in frontier AI's commercial future and public confidence in its day-to-day reliability are now moving on separate tracks.

Why It Matters: A potential $45 billion compute deal underscores that frontier AI competition is increasingly about infrastructure access—and that investor confidence remains high despite mounting safety concerns.