🔍 Read the full analysis: The Cost Of Leaving Claude: Lessons From Meta And Microsoft on ThorstenMeyerAI.com
Get school and study supplies delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
TL;DR
The Information reported on Oct. 5 that Meta and Microsoft are reducing some employees’ use of Anthropic’s Claude tools and directing them to alternatives. The reported reasons include costs and the companies’ existing tools; the report does not establish that Claude performed worse, and both companies reportedly continue to use Anthropic technology in other ways.
Meta and Microsoft are redirecting some employees from Anthropic’s Claude tools to in-house or rival products, according to a report by The Information on Oct. 5. The reported shifts point to cost control and available alternatives, not a stated finding that Claude performs worse, and do not mean either company has ended its use of Anthropic technology.
The Information reported that Meta’s Claude Code users fell from about 60,000 earlier this year to about 30,000. During the same period, Meta’s own coding products gained internal users: the report put MetaCode above 30,000 users and Muse Code above 6,000. The figures describe employee use, not adoption by Meta’s customers.
Microsoft had reportedly projected more than $1 billion a year in internal spending on Anthropic technology, including Claude Code, Claude models in Copilot and Claude Mythos. The report says Microsoft cut that projection by more than a third and is steering employees toward GitHub Copilot and OpenAI models. That is a change to a reported projection; it does not, by itself, establish the company’s final spending or the exact amount already saved.
The supplied account of the reporting says Microsoft still uses Anthropic models for some customer-facing Copilot features, while customer spending on Claude through Microsoft platforms is reportedly growing. It also attributes the internal pullback to token costs, tighter spending controls and a preference for tools the companies own or back. Neither company is reported to have said Claude’s quality was the reason.
Meta and Microsoft pulled back from Claude. Here’s what switching actually costs.
The Information reports both companies steering their own employees away from Claude. Read as a verdict on Claude, it misleads. Read as a demonstration of switching — and who can afford it — it’s the most useful enterprise-AI signal this month.
Staff steered to GitHub Copilot and OpenAI models; stricter token budgets. One unconfirmed report: some team budgets ~$100k → ~$10k/month.
Microsoft reportedly still spends heavily on Claude for customer-facing Copilot — and that spending is reported to be growing.
Reported drivers: rising token costs and owned alternatives. Neither company is reported to have called Claude worse.
Meta builds coding tools; Microsoft owns Copilot and backs OpenAI. This is ordinary vertical integration.
Keep a second vendor live on real work.
A few hundred tasks with pass criteria.
Logic, prompts, tools in your layer.
Tokens are the cheap half.
Know what you’d rebuild.
On the evidence reported, Meta and Microsoft didn’t reject Claude. They brought spending in-house where they could and kept buying where they couldn’t — Microsoft remains a large Anthropic customer for the products it sells. The signal is the mechanism: the most sophisticated buyers treat models as interchangeable suppliers behind a layer they control.Meta could halve its Claude usage because it had built somewhere else to go. Build somewhere else to go.
Switching Requires More Than a New Model
The shifts show how large buyers can respond when AI bills rise: route work to another provider or internal tool instead of relying on one model for every task. Meta and Microsoft have products and engineering teams ready to support that move. Their choices therefore illustrate a procurement option, not proof that the same move would pay off for a company without comparable substitutes.
Changing tools can bring costs that do not appear in a token price. Companies may need to rerun evaluations, adapt prompts and software integrations, and give employees time to learn a different workflow. If a replacement performs less well on a company’s specific tasks, extra review and rework can erode savings. The source material identifies these as potential switching costs; it does not quantify them for either company.
The practical point for other buyers is to measure cost per accepted result, not just model usage, and to prepare a safe alternative before a price or policy change forces a decision. A second model in limited production and a set of representative tests can make a future comparison more concrete. They cannot guarantee that switching will be cheaper or preserve output quality.
As an affiliate, we earn on qualifying purchases.
Both Companies Have Alternatives
This report concerns internal employee use, a narrower question than whether customers can buy or use Claude. Meta develops its own AI models and coding tools. Microsoft owns GitHub Copilot and has a major relationship with OpenAI. That makes both companies unusual buyers: each can direct some work toward products it owns, develops or backs.
The reported figures also cover different measures. Meta’s numbers refer to employees using coding tools, while Microsoft’s figure is a projected annual internal spend across several Anthropic products. They should not be treated as directly comparable measures of use or savings. The supplied account says Microsoft’s internal budget controls also tightened; one account put some monthly team budgets at about $10,000, down from around $100,000. That figure is based on a single report and should not be generalized across Microsoft.
The report’s account separates these internal decisions from ongoing customer-facing activity. On the information provided, neither company has announced a complete break with Anthropic, and the reported employee shifts do not establish a broad decline in customer demand for Claude.
The Scale of the Shift Is Unclear
The account does not provide a direct statement from Meta, Microsoft or Anthropic explaining the reported decisions. It is therefore unclear how the companies define active use, how many employees are affected beyond the figures cited, or how quickly the changes took place. The reported user counts and spending projection have not been independently verified within the supplied material.
It is also unclear how much of the projected reduction will become actual savings, what share of workloads remains on Anthropic models, and how the alternative tools compare on quality, reliability and total cost for each company’s work. The report gives no task-level performance data. The reasons attributed to the moves are costs and existing alternatives, but the available information does not rule out other considerations.
Watch Actual Spend and Usage
The next useful evidence would be updated figures showing actual internal spending and employee use, rather than forecasts alone, alongside clarification about which workloads remain on Claude. Statements from the companies could also explain whether the moves are broad policy changes or targeted adjustments to coding-tool budgets.
For enterprise buyers, the relevant follow-up is whether the alternatives deliver comparable results after integration and review costs are counted. Until those details are available, the report supports a limited conclusion: two large companies with their own credible options are shifting some internal work away from Claude, while the full financial and operational effects remain unknown.
Key Questions
Have Meta and Microsoft stopped using Claude?
No complete stop has been reported. The report concerns internal employee use. The supplied account says Microsoft continues to use Anthropic models for some customer-facing Copilot features.
Why are the companies reportedly shifting employees?
The reported reasons include token costs, tighter spending controls and available alternatives. Neither company is reported to have said Claude performed worse.
How much is Microsoft expected to save?
The Information reportedly said Microsoft cut a projected annual internal Anthropic spend of more than $1 billion by more than a third. That is a change to a projection, not a confirmed figure for realized savings.
Does this mean other companies should switch from Claude?
Not necessarily. Meta and Microsoft have their own tools and engineering capacity. Other organizations would need to compare model performance and account for integration, evaluation, employee training, review and rework costs.
Source: ThorstenMeyerAI.com
Evergreen bestsellers Picks
bestsellers
As an affiliate, we earn on qualifying purchases.
