🔍 Read the full analysis: Claude’s Place In AI: What Meta And Microsoft’s Pullback Means For Switching on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
The Information reported on Oct. 5 that Meta has reduced employee use of Claude Code and Microsoft has lowered its internal spending projection for Anthropic technology, while directing staff toward alternatives. The reported changes concern internal use, not a broad end to Claude access or a confirmed decline in model quality. They show how switching depends on having substitutes and the ability to absorb the engineering and productivity costs.
Meta and Microsoft are steering some employees away from Anthropic’s Claude tools and toward alternatives they own or already use, according to a report by The Information on Oct. 5. The reported moves concern the companies’ internal use, not a decision to end Claude access for customers, and point to the practical challenge of switching AI systems: savings depend on having a credible replacement and the capacity to move work to it.
Meta reportedly reduced the number of employees using Claude Code from about 60,000 earlier this year to about 30,000. The report says staff are being directed toward Meta’s internal coding tools: MetaCode, which has more than 30,000 internal users, and Muse Code, which has more than 6,000. These figures describe reported internal usage; they do not establish how much coding work moved or whether the tools deliver comparable results.
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 Information reported that the projection has since been cut by more than a third, with employees steered toward GitHub Copilot and OpenAI models. A separate detail in the source account says some monthly team budgets may have fallen from about $100,000 to about $10,000; that figure is attributed to a single report and has not been independently established here.
The reported reasons include rising token costs, tighter spending controls and investment in in-house tools. Neither company is reported as saying Claude performed worse. Microsoft is also reported to continue spending on Anthropic models for customer-facing Copilot features, while customer spending on Claude through Microsoft platforms is said to be growing. The account does not describe a general cutoff of Claude access.
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 Substitute
The moves matter because they show that AI purchasing choices can shift inside large companies when costs rise or another tool is available. But the headline usage changes do not mean every buyer can make the same move at the same speed. Meta and Microsoft have substantial engineering resources and alternatives already in place; many companies do not.
Switching models can require teams to rerun evaluations, adapt prompts and tool integrations, and retrain staff on a different system. It can also interrupt work while people adjust. For coding agents, performance depends partly on connections to editors, repositories and team workflows. A replacement that looks cheaper per token may still cost more if it leads to additional review, rework or errors.
The companies’ reported choices support a case for making systems portable before a supplier change is needed, rather than assuming a switch is effortless. That does not prove that every organization should use multiple models: maintaining integrations and evaluating alternatives also takes time and money. Buyers need to compare the full cost and quality of accepted work, not just a model’s listed price or token bill.
As an affiliate, we earn on qualifying purchases.
Internal Tools, Not a Customer Exit
The report concerns Meta’s and Microsoft’s own employees. That distinction limits what can be concluded about Anthropic’s wider business. Microsoft reportedly continues to use Anthropic models in customer-facing Copilot features, and the source account says customer purchases of Claude through Microsoft platforms are growing. Internal procurement changes can coexist with continued sales or use elsewhere.
Both companies also have reasons beyond supplier cost to promote alternatives. Meta develops its own models and coding products; Microsoft owns GitHub Copilot and is a major backer of OpenAI. The report therefore describes decisions by buyers that also have competing tools and commercial interests. It is not, by itself, an independent comparison of Claude’s quality against those alternatives.
The source account cites a separate SemiAnalysis finding that AI subscription limits can change without clear notice and that list-price reductions may lower what a subscription provides. That observation adds context to the broader issue of changing AI product terms, but it does not establish the terms of Meta’s or Microsoft’s Anthropic arrangements. No contract details are provided in the report summary.
As an affiliate, we earn on qualifying purchases.
Usage Figures Leave Key Gaps
The precise terms and timing of the reported changes are not clear from the material provided. It does not include statements from Meta, Microsoft or Anthropic confirming the figures, explaining the full scope of the cuts, or setting out how quickly employees are expected to move. The reported user counts also do not reveal frequency of use, workload volume or whether people retain access to Claude.
It is also unknown how the replacement tools compare on the companies’ own tasks, or whether the changes reflect measured quality, cost targets, product strategy or a combination. The report’s stated drivers are cost and internal alternatives; that is different from evidence that Claude’s performance has declined. The effect on Anthropic revenue is not quantified, and the internal spending projection should not be treated as actual annual spending.
More broadly, the available details cannot show whether similar savings are realistic for smaller buyers. The engineering effort, evaluation quality and productivity effects of a move will vary by organization and workload. Any claim that switching will pay off for a particular company requires its own cost and performance data.
AI development environment software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Watch for Confirmed Spending Details
The next useful evidence would be direct confirmation from the companies about internal usage, spending and the scope of any employee guidance. Further detail on Microsoft’s Anthropic use in customer products would help distinguish internal substitution from changes to customer-facing offerings. No specific public milestone or completion date for the reported shifts is given.
For organizations weighing their own AI choices, the immediate task is to measure representative work across available tools, including review time and rework alongside token costs. Maintaining a tested alternative and keeping prompts, evaluations and business logic portable may lower future switching friction, but the cost of doing so should also be counted. The reported moves show what two well-resourced companies are doing; they do not settle which model is best for other buyers.
As an affiliate, we earn on qualifying purchases.
Key Questions
Are Meta and Microsoft ending their use of Claude?
No such broad cutoff is reported. The account describes changes to internal employee use. Microsoft is also reported to continue using Anthropic models in customer-facing Copilot features.
Why are the companies reportedly shifting some work?
The reported reasons are rising token costs, tighter spending controls and available internal or existing alternatives. Neither company is reported to have said Claude performed worse.
How much did Meta’s reported Claude Code usage change?
The Information reported that the number of Meta employees using Claude Code fell from about 60,000 earlier this year to about 30,000. The figures do not show how much work those employees completed with the tool.
Does this mean other companies should switch models?
Not necessarily. Switching can involve new evaluations, integration work, staff adjustment and possible review or rework costs. Buyers need to compare performance and total cost on their own tasks before deciding.
Source: ThorstenMeyerAI.com
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
