🔍 Read the full analysis: A Closer Look At Switching Costs After Meta And Microsoft Pulled Back From Claude on ThorstenMeyerAI.com
Get hardware and tech essentials 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 roughly halved employee use of Claude Code and Microsoft cut its internal Anthropic spending projection by more than a third. The reported moves reflect cost controls and available in-house or rival tools; they do not establish that either company has stopped using Claude or judged it inferior.
Meta and Microsoft have reportedly reduced internal use or projected spending on Anthropic’s AI tools, directing employees toward alternatives they own or already use, according to The Information on Oct. 5. The reported changes point to cost controls and switching options inside two large technology companies, but do not establish that either company has ended its use of Claude or found it performs worse.
Meta reportedly reduced the number of employees using Claude Code from about 60,000 earlier this year to about 30,000. The report says the company has steered engineers toward its own coding tools: MetaCode, which has more than 30,000 internal users, and Muse Code, with more than 6,000. These are figures attributed to the reporting; Meta’s internal usage figures and the precise timing of the shift were not independently detailed in the supplied account.
Microsoft reportedly had projected more than $1 billion a year in internal spending on Anthropic technology, including Claude Code, Claude models in Copilot and Claude Mythos. It has since cut that projection by more than a third and directed employees toward GitHub Copilot and OpenAI models, according to The Information. The report also describes stricter token budgets. One account said some monthly team budgets fell from roughly $100,000 to roughly $10,000; that specific figure comes from a single report and should not be treated as a company-wide policy.
The reported pullback concerns the companies’ own employees, not a complete withdrawal of Anthropic technology from customer-facing products. The source account says Microsoft continues to spend on Anthropic models for Copilot features and that customer spending on Claude through Microsoft platforms is growing. Neither company is reported to have said that Claude’s quality drove the internal changes.
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 Large Buyers Can Switch
The development matters because it shows how cost, control and alternatives can shape enterprise AI purchasing even when a company is not publicly rejecting a supplier’s product. A buyer that already operates another model or coding assistant can redirect work when budgets change. That option may give large companies leverage over vendors and limit dependence on one provider’s prices, availability or terms.
But the reported moves are not a simple template for other businesses. Meta and Microsoft have their own tools, substantial engineering capacity and multiple products to draw on. For a company without those resources, changing models can involve fresh testing, prompt and integration work, employee retraining, and uncertainty about whether a replacement performs well on its actual tasks. Those costs may offset savings, and the reporting does not quantify them for either company.
The practical point is not that every organization should leave Claude or adopt a particular competitor. It is that the ability to switch depends on preparation. Keeping model-independent application logic, testing more than one provider on real work and measuring accepted results can make future choices more informed. Those measures have costs too; their value depends on a company’s workload, risk tolerance and expected savings.
As an affiliate, we earn on qualifying purchases.
The Report’s Key Limits
The account describes internal employee use, not a broad customer migration. It also distinguishes Microsoft’s internal spending projection from its reported use of Anthropic models in customer-facing Copilot features. Those details matter: a company can cut the amount its own staff use a vendor’s tools while continuing to sell products that rely on that vendor’s models.
The companies named in the report also have potential substitutes and business interests in the market. Meta develops its own models and coding tools; Microsoft owns GitHub Copilot and has a major relationship with OpenAI. Their reported decisions can therefore reflect in-house capacity and product strategy alongside price. The available information does not show that an ordinary enterprise could make the same change at the same cost or pace.
For teams weighing model options, the underlying work can extend beyond changing an API. They may need to rerun evaluations, adapt prompts and tool definitions, rebuild integrations, and account for changes in context caching and review effort. These are plausible sources of switching cost described in the source material, not documented cost totals for Meta or Microsoft.
“Meta reduced employee use of Claude Code from about 60,000 earlier this year to about 30,000.”
— The Information, as described in the supplied source
As an affiliate, we earn on qualifying purchases.
What the Usage Figures Show
The supplied account does not include statements from Meta, Microsoft or Anthropic confirming the reported numbers or explaining the timing and scope of the changes. It is therefore unclear how many employees still use Claude Code, how the internal spending projection compares with actual spending, or whether the reported budgets apply broadly or only to particular teams.
It is also not clear how much the changes save after accounting for engineering work, employee productivity, review and rework, or differences in model performance. The report, as summarized here, gives cost and in-house alternatives as drivers, but it does not supply comparable task-quality results or a complete cost analysis. The account says Microsoft continues to use Anthropic models in customer-facing Copilot features; the scale and terms of that use are not specified.
As an affiliate, we earn on qualifying purchases.
Further Details Could Clarify Scope
The next useful developments would be direct responses from the companies or additional reporting that clarifies the timing, scope and spending figures. More information could establish whether the shifts affect specific engineering teams or wider internal deployments, and whether Microsoft’s lower projection changes its customer-facing use of Anthropic models.
For enterprise buyers, the reported moves may prompt closer scrutiny of how model contracts, usage budgets and alternatives are managed. Any decision to switch still depends on measured performance and the full cost of moving a workflow—not token prices alone. Until more detail is available, the report supports a narrower conclusion: these companies are directing some internal work toward available alternatives, while the extent and consequences of the changes remain uncertain.
As an affiliate, we earn on qualifying purchases.
Key Questions
Have Meta and Microsoft stopped using Claude?
No such complete halt is established. The report concerns reduced internal use or projected spending. The supplied account says Microsoft continues to use Anthropic models for customer-facing Copilot features.
Why did the companies reportedly reduce internal use?
The reported reasons are rising token costs, tighter budgets and available in-house or rival tools. Neither company is reported to have said that Claude performed worse.
How much did Microsoft reportedly cut its projection?
Microsoft reportedly cut its projected internal spending on Anthropic technology by more than a third from a projection of more than $1 billion a year. The source account does not provide a confirmed actual-spending total.
Does this mean other companies should switch AI models?
No. The report does not establish that switching is right for other buyers. A change can bring testing, engineering and retraining costs, so organizations would need to compare total costs and task performance for their own workloads.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
