What's Happening?
Meta Platforms has reportedly reduced the number of its internal developers using Anthropic's Claude Code by half, from approximately 60,000 to 30,000. This move signifies Meta's strategic shift towards its proprietary coding tools, MetaCode and Muse
Code. The reduction in usage was highlighted by investor Paul Kedrosky on 'The Ezra Klein Show,' and further reported by The Information. Microsoft has also reportedly cut its projected internal spending on Claude by over a third. This development comes as Meta, previously one of Anthropic's largest customers, now possesses its own frontier AI model. Despite the significant reduction, Meta still spent over $105 million on Claude Code in a recent 28-day period, indicating the substantial scale of its prior engagement with Anthropic's services. This shift underscores a broader trend among major tech companies to develop and utilize their own AI capabilities.
Why It's Important?
This reduction in Meta's reliance on Anthropic's Claude Code poses a significant challenge for Anthropic, a leading AI developer. Anthropic's business model is heavily dependent on a small number of large corporate clients, with investor Paul Kedrosky suggesting that as few as eight companies might account for 80% of its revenue, primarily for code generation. Meta's decision to transition to its own AI tools directly impacts Anthropic's revenue streams and growth trajectory. While Meta's Claude Code spending in 2025 was estimated to be around 10% of Anthropic's revenue, and currently represents about 2% of Anthropic's latest quarterly pace, the loss of such a major client can signal a broader trend. This move by Meta, coupled with similar actions by Microsoft, indicates a growing self-sufficiency among tech giants in AI development, potentially leading to increased competition and reduced demand for third-party AI services.
What's Next?
The immediate consequence for Anthropic is a potential slowdown in its revenue growth, as indicated by Bloomberg's report of decelerating top-line growth in its third quarter. As major clients like Meta and Microsoft reduce their reliance on external AI tools, Anthropic will need to diversify its customer base or innovate its offerings to maintain its market position. Other large companies that currently rely on Anthropic's services might follow Meta's lead, further intensifying the competitive landscape in the AI sector. For Meta, this transition is expected to lead to greater control over its AI infrastructure and potentially cost savings in the long run, although the immediate financial impact is relatively small compared to its overall capital expenditure. The broader AI industry may see an acceleration in the development of in-house AI models by large corporations, potentially reshaping the market for AI services and tools.
Beyond the Headlines
This development highlights a critical juncture in the evolution of the AI industry, where the initial phase of widespread adoption of third-party AI models is giving way to a new era of proprietary AI development by tech giants. The strategic implications extend beyond mere financial figures, touching upon issues of data security, intellectual property, and competitive advantage. Companies like Meta are seeking to internalize AI capabilities not just for cost efficiency but also to tailor AI models more precisely to their specific needs and to maintain a competitive edge in innovation. This trend could lead to a more fragmented AI ecosystem, where large players operate with their own sophisticated AI stacks, potentially creating barriers for smaller AI firms that rely on providing foundational models. The long-term impact could be a shift in power dynamics within the tech industry, with greater consolidation of AI expertise and infrastructure among a few dominant players.
