🔍 Read the full analysis: Meta And Microsoft Step Back From Claude: The Economics Of Switching on ThorstenMeyerAI.com
Get the little things that make your day 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 reduced employee use of Claude Code and Microsoft cut its forecast for internal Anthropic spending by more than a third. The reported moves concern the companies’ own workforces, not an end to Claude access or a broad rejection of Anthropic’s models. Both companies have alternatives in place, a condition that may make switching harder for other businesses.
Meta and Microsoft have reportedly reduced some employees’ use of Anthropic’s Claude tools, citing internal costs and available alternatives, according to a report by The Information on Oct. 5. The reported changes concern internal use, not a company-wide withdrawal of Claude from customer-facing products, and illustrate how access to substitutes can affect the cost of switching AI providers.
Meta reportedly reduced the number of employees using Claude Code from about 60,000 earlier this year to about 30,000. The company has directed staff toward its own coding tools: MetaCode, which the source report says has more than 30,000 internal users, and Muse Code, with more than 6,000.
Microsoft had reportedly projected annual internal spending of more than $1 billion on Anthropic technology, including Claude Code, Claude models used in Copilot, and Claude Mythos. The report says Microsoft has cut that forecast by more than a third and is steering employees toward GitHub Copilot and OpenAI models. The source material also says Microsoft continues to use Anthropic models for customer-facing Copilot features, and that customer spending on Claude through Microsoft platforms is growing.
The reported reasons are cost controls and competing tools, not a stated finding that Claude performs worse. Neither company is reported as saying it has ended access to Claude. A separate detail in the source material says some Microsoft team token budgets fell from around $100,000 a month to around $10,000; that figure is attributed to a single account and should not be treated as a company-wide policy.
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 Depends on Existing Alternatives
The reported changes show how large AI buyers can respond when a supplier’s costs become harder to justify: shift some work to another model or tool rather than rely on one provider. But that option is easier for Meta and Microsoft because they already have alternatives in use and substantial engineering capacity. Their decisions do not establish that the same move would save money for a company without comparable tools.
For other businesses, the price of switching can include rechecking model quality on real tasks, adapting prompts and software integrations, and allowing time for staff to learn a new system. Moving providers can also change caching costs for workloads that repeatedly use stored context. And if the replacement is less effective for a company’s work, additional review and rework may reduce or erase savings. These costs are not necessarily visible in a model’s listed token price.
The practical implication is not that companies should leave Claude, or that they should stay. It is that price comparisons alone are incomplete. Buyers need to compare the cost of producing an accepted result, including human review, errors and engineering work, and to know how much it would cost to move a real workflow elsewhere.
As an affiliate, we earn on qualifying purchases.
Internal Use, Not a Customer Exit
The distinction between employee use and customer products is central to interpreting the report. Meta and Microsoft are large organizations that also build or back competing AI products. Meta develops its own models and coding tools; Microsoft owns GitHub Copilot and is OpenAI’s largest backer, according to the source material. Their internal decisions therefore reflect both purchasing choices and the availability of products in which they have a direct stake.
Microsoft’s reported reduction in its internal Anthropic spending forecast does not, on the information provided, mean it has stopped using Claude in Copilot for customers. The report also does not say that either company has removed Claude from all employee workflows. The headline framing of two companies stepping back from Claude should not be read as evidence of a general product-quality verdict.
The report’s figures point to a potential financial incentive at Microsoft, but they remain estimates about internal spending. A reduction of more than a third from a forecast above $1 billion would amount to a large projected cut, yet the source material does not provide a final realized spending total or a precise period over which the change will take effect.
“Meta reportedly reduced Claude Code use among employees from about 60,000 earlier this year to about 30,000.”
— The Information, in reporting dated Oct. 5
As an affiliate, we earn on qualifying purchases.
The Scale of the Cuts Is Unclear
The available account does not specify the full timeline for the changes, how many teams or workflows are affected, or how the companies measured the performance of the alternatives. It also does not establish whether the reductions are permanent, whether employees can still use Claude for particular tasks, or how much of the forecast reduction will appear in actual spending.
There is no reported statement from either company that Claude underperformed on its tasks. The reasons described are cost, tighter spending controls and a preference for tools the companies own or support. The reported figures also do not show whether switching brought changes in productivity, review time or quality. Without those measures, the financial effect of the moves cannot be fully assessed.
As an affiliate, we earn on qualifying purchases.
Watch Actual Spending and Results
The next useful evidence would be clearer figures on realized spending, internal user numbers and which work has moved to alternative tools. Performance comparisons would also help establish whether the changes reduced total costs or shifted expenses into engineering, review and rework.
For businesses weighing similar decisions, the reported moves underline the value of preparing before a price change forces a choice: test more than one model on representative tasks, retain a way to compare results, and track both token costs and the human effort needed to approve output. Those steps do not guarantee that switching will pay off, but they can make the trade-offs more measurable.
As an affiliate, we earn on qualifying purchases.
Key Questions
Have Meta and Microsoft stopped using Claude?
The report describes reductions in some internal use and spending forecasts, not a complete end to Claude access. It says Microsoft continues to use Anthropic models for some customer-facing Copilot features.
Why are the companies reportedly reducing internal use?
The reported reasons are rising costs, tighter spending controls and the availability of alternatives, including Meta’s own coding tools and Microsoft’s GitHub Copilot and OpenAI models. Neither company is reported as saying Claude performed worse.
How much did Microsoft reportedly cut its forecast?
Microsoft reportedly reduced its projection for internal annual spending on Anthropic technology by more than a third from a forecast above $1 billion. The source material does not give a final actual spending total.
Does this mean other companies should switch AI providers?
Not necessarily. Switching can require new evaluations, engineering changes and employee retraining, and it may affect output quality and review costs. Meta and Microsoft have alternatives already in place, which may make a move easier for them than for many other organizations.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
