Meta internally projected it could spend up to $10 billion on Anthropic's artificial intelligence services, according to a report published by The New York Times. The figure, drawn from internal Meta documents or discussions, offers a striking window into how much the social media giant values access to cutting-edge AI models it does not build itself. It also adds new detail to a story that has been developing for months around the relationship between Meta and Anthropic.
What the Projection Means
A $10 billion figure is not a signed contract. It is a projection, meaning Meta's teams modeled a scenario in which that level of spending on Anthropic's AI could be justified. Still, internal projections at companies the size of Meta are rarely speculative fiction. They reflect genuine demand analysis, capacity planning, and strategic intent. Reports earlier this year indicated that Meta was in talks to lease $10 billion in computing power to Anthropic, suggesting the two companies have been exploring a deeper commercial relationship from multiple angles.
Key Facts
- Meta internally projected potential spending of up to $10 billion on Anthropic's AI
- The figure was reported by The New York Times
- Earlier reports described separate talks about a $10B computing arrangement between the two companies
- Anthropic has secured major investment from Google and Amazon in recent months
- Meta develops its own Llama models but has explored external AI partnerships
The timing matters. Anthropic has been pulling in capital at a pace that few AI companies can match. Google committed up to $40 billion to Anthropic in what stands as the largest AI investment on record, and Amazon has made similarly large commitments on the compute side. If Meta were to become a significant customer or partner, it would further cement Anthropic's position as one of the central players in the enterprise AI market, even as Meta continues to develop its own open-weight Llama models in parallel.
Internal projections of this scale suggest Meta views Anthropic's models as genuinely useful for products or infrastructure it cannot build fast enough internally.Industry analysis based on NYT reporting
A Competitive Market for AI Access
The broader context here is a technology industry engaged in something close to a land rush for AI capacity. Companies that once prided themselves on building everything in-house are now writing large checks to access models from specialized AI labs. Anthropic, with its Claude model lineup, has positioned itself as a credible alternative to OpenAI for enterprise customers who want reliability, safety focus, and strong performance. Questions about whether that growth can be sustained at current spending levels have surfaced, with some analysts noting that AI sticker shock could slow Anthropic's expansion if customers push back on pricing.
For now, the reported Meta projection suggests that at least one of the world's largest technology companies sees enough value in Anthropic's offering to model nine or ten-figure spending. Whether those projections translate into signed agreements, and on what terms, remains to be seen. Anthropic has not publicly confirmed the details of any arrangement with Meta, and Meta has not issued a formal statement on the NYT report. What is clear is that access to frontier AI models has become a budget line item that major corporations are taking seriously, and Anthropic sits at the center of that calculus.
