🔍 Read the full analysis: Meta And Microsoft Pull Back From Claude, Raising Questions About AI Costs on ThorstenMeyerAI.com
Get smart everyday buys 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
Meta and Microsoft have reportedly reduced internal employee use of Anthropic’s Claude tools and directed more work to alternatives they own or back. The reported moves concern internal use, with cost and spending controls cited; they do not show that either company has ended Claude access or that Claude performed worse.
Meta and Microsoft have reportedly scaled back employee use of Anthropic’s Claude tools, steering staff toward products they own or back as internal AI costs come under pressure. The Information reported the changes on Oct. 5, describing a reduction in Meta’s Claude Code users and a lower Microsoft forecast for internal Anthropic spending; neither move, as reported, amounts to a shutdown of Claude access or a verdict on its quality.
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 products: MetaCode, which the source report says has more than 30,000 internal users, and Muse Code, with more than 6,000. These figures describe internal adoption, not use 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 company has since cut that projection by more than a third, according to the report, and is directing employees toward GitHub Copilot and OpenAI models. The report also says Microsoft continues to use Anthropic models in customer-facing Copilot features and that customer spending on Claude through Microsoft platforms is growing.
The reported reasons for the internal changes are rising token costs, tighter spending controls and alternatives developed or backed by the companies. Neither company is reported to have said that Claude performed worse. The source material also describes stricter Microsoft token budgets; one account puts some monthly team budgets at roughly $10,000, down from about $100,000. That budget detail comes from a single report and should not be treated as a company-wide figure.
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.
Why Internal AI Budgets Are Shifting
The reported changes show how major companies may respond when the cost of using AI tools grows: shift work to another provider or an internal product, rather than rely on one supplier for every task. For buyers, the relevant question is not only model price, but whether an alternative can handle existing work without adding more engineering, review or operational cost.
Meta and Microsoft have an advantage that most organizations do not: they already have competing tools in place. Meta has its own coding products, while Microsoft can steer employees toward GitHub Copilot and OpenAI models. That makes switching more feasible for them than for a company whose applications, staff workflows and evaluations depend on one model.
Moving tools still carries costs that may not appear on a token bill. Teams may need to repeat evaluations, adjust prompts and tool connections, and give employees time to learn a different system. Changes can also affect caching economics and the amount of human review required. If a replacement model produces weaker results on a company’s specific tasks, the cost may show up as rework or errors rather than a higher invoice. The report gives no measured comparison of those costs or of output quality, so the savings cannot be read as proof that one tool is better overall.
The implication for other AI buyers is that having the ability to switch can be valuable, but only if the alternative has been tested on real work. A second model or provider may reduce dependence on one vendor, while poorly planned migration can erase savings through integration work and lost productivity. Cost per token alone is an incomplete measure; organizations also need to track the effort and quality involved in completing a task.
As an affiliate, we earn on qualifying purchases.
Substitutes Make Switching Possible
The report concerns how Meta and Microsoft use AI tools inside their own organizations. That distinction matters because internal usage and customer-facing services are separate: Microsoft’s reported cut to its internal spending projection does not mean it has stopped using Anthropic models in Copilot features or that customers have lost access through Microsoft platforms.
Both companies also have interests in alternatives to Claude. Meta develops its own models and coding products. Microsoft owns GitHub Copilot and is a major backer of OpenAI. That makes their decisions different from those of a buyer without comparable products already deployed. Their reported shifts can reflect efforts to control costs and build around existing tools; they do not, by themselves, establish that Claude is unsuitable for other customers.
The source material points to a broader purchasing issue: AI subscription limits and token costs can change, and switching providers can carry hidden expenses. Those factors make portability and evaluation important, but no independent cost comparison or complete account of the companies’ migration work is provided here.
As an affiliate, we earn on qualifying purchases.
What the Report Does Not Establish
The exact timing, scope and final savings remain unclear. The available account does not specify which employees or workflows were included in each change, how much Anthropic technology Microsoft still uses internally, or whether the reported spending reduction became actual spending rather than a revised projection.
The figures on internal users and budgets are attributed to reporting, and the source material does not provide direct statements from Meta, Microsoft or Anthropic explaining the decisions. It also does not give comparative performance data for Claude and the replacement tools, or quantify the engineering and productivity costs of moving workloads. No reported quality finding should be inferred from a cost-driven shift.
It is also unclear whether the changes apply to all internal teams or only selected uses. The reported growth in customer Claude spending through Microsoft platforms and continued customer-facing use are distinct from employee adoption and do not reveal how either may change later.
As an affiliate, we earn on qualifying purchases.
The Next Test Is Measured Performance
The next useful indicators are whether Meta and Microsoft maintain or expand the reported shifts, and whether either company publishes more detail about internal spending, usage or results. Further reporting or company statements could clarify the number of teams affected, the pace of the move and how customer-facing use of Claude is being handled.
For other organizations, the practical next step is to compare models on representative tasks before moving important workloads. That means setting clear pass criteria, checking output quality and review time, and accounting for integration and staff costs alongside token charges. A small, tested alternative can show whether switching is workable; a headline about two large buyers cannot supply that answer for every company.
As an affiliate, we earn on qualifying purchases.
Key Questions
Have Meta and Microsoft stopped using Claude?
No such full stop is reported. The account describes reduced or lower projected internal use. It also says Microsoft continues to use Anthropic models in customer-facing Copilot features.
Why are the companies reportedly shifting employee use?
The reported reasons are rising token costs, tighter spending controls and the availability of alternatives. The source material does not say either company found Claude performed worse.
How much did Microsoft reduce its projected Anthropic spending?
Microsoft reportedly cut a projection of more than $1 billion a year in internal spending by more than a third. The report does not establish the final amount actually spent.
Does this mean customers are losing access to Claude?
The reported changes concern internal employee use. The account says Microsoft continues to use Anthropic models for customer-facing Copilot features and that customer spending on Claude through Microsoft platforms is growing.
What remains unknown about the reported moves?
The full scope, timing and realized savings are not clear. The source material also provides no comparative performance results showing how Claude and the alternatives perform on the companies’ work.
Source: ThorstenMeyerAI.com
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
