What is Driving Meta and Microsoft to Reduce Claude Usage?
Meta and Microsoft, two of the largest tech companies globally, are cutting back on their internal use of Anthropic's Claude AI platform. The primary driver behind this move appears to be the high expense associated with using Claude at a large scale. Microsoft reportedly faced internal costs approaching $1 billion this year alone, making reliance on an external AI system economically challenging.
Both companies have developed their own proprietary AI tools—Meta with MetaCode and Microsoft with Copilot. Switching to these internally controlled platforms helps reduce costs and grants tighter integration with their existing workflows. For instance, Meta has halved its Claude user seats from 60,000 to 30,000, while MetaCode usage has grown beyond 30,000 users, indicating a clear transition to its homegrown solution.
How Do Security Concerns Impact Anthropic's Reputation?
Beyond cost, Anthropic as a company is experiencing reputational and potentially operational challenges after being labeled a national security supply chain risk by the US Department of Defense. This political designation may influence other customers' willingness to engage with Anthropic, creating a ripple effect on contracts and future business.
The designation could increase uncertainty for enterprises who prioritize security and regulatory compliance. Anthropic's prospectus itself warns about possible material revenue losses from these developments, especially given that Microsoft and Meta represent significant portions of its client base.
What Are the Broader Implications for AI Usage in Big Tech?
This shift reflects a wider economic and strategic trend: companies increasingly prefer in-house or more affordable AI options over third-party providers, even if the alternatives have slightly less advanced capabilities. Smaller US organizations have similarly gravitated towards cheaper alternatives, including Chinese AI models, to manage budgets effectively.
High-volume AI usage can generate enormous costs, underlining the importance of balancing performance against expenses and operational control. Using internal AI tools facilitates spending limits and monitoring that private providers may not offer.
Practical Takeaways for AI Users and Buyers
For organizations considering AI platforms, this situation highlights the importance of evaluating total cost of ownership, including ongoing usage expenses and potential geopolitical or regulatory risks linked to vendors. Developing internal AI capabilities or negotiating stricter usage controls with external providers may improve cost predictability.
Users should not assume third-party AI platforms will be a permanent solution, especially when the provider faces reputational challenges or governmental scrutiny. Staying informed about vendor stability and industry trends will help ensure AI investments remain sustainable and aligned with organizational needs.