🔍 Read the full analysis: How A Claude Switch Could Affect Your AI Budget on ThorstenMeyerAI.com
Get privacy and security gear 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 redirecting some employees from Anthropic’s Claude tools toward alternatives they own or use. The reported changes concern internal use, not an end to Claude access or a confirmed judgment about quality. For other companies, switching can lower model bills but also brings engineering, evaluation, integration and productivity costs.
Meta and Microsoft are steering some employees away from Anthropic’s Claude coding tools and toward alternatives, according to a report by The Information on Oct. 5. The reported moves point to a potential way for large AI buyers to manage spending, but they do not establish that Claude has been dropped, that customer access is ending, or that the companies found it inferior.
The Information reported that Meta reduced the number of employees using Claude Code from about 60,000 earlier this year to about 30,000. The report said Meta has directed staff toward its internal coding tools: MetaCode, which has more than 30,000 internal users, and Muse Code, with more than 6,000.
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. According to the report, Microsoft cut that projection by more than a third and directed employees toward GitHub Copilot and OpenAI models. One reported account also put some monthly team budgets at about $10,000, down from roughly $100,000; that figure comes from a single report and should not be treated as a company-wide budget rule.
The reporting describes internal employee use, not a general withdrawal of Claude from either company’s customer-facing products. Microsoft reportedly continues to spend on Anthropic models for Copilot features, while customer spending on Claude through Microsoft platforms is reported to be growing. The reported drivers of the internal changes are cost controls and in-house alternatives, not a public finding that Claude performs worse.
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 Switching Changes the Budget
For businesses buying AI, the report highlights a distinction between a model’s listed price and the total cost of moving work. A lower token bill can be offset by staff time spent adapting prompts, rebuilding integrations, testing workflows and reviewing changed outputs. If a replacement performs less well on a company’s tasks, additional corrections and quality checks can add costs that do not appear on an API invoice.
Large technology companies can absorb these costs more readily when they already have alternatives deployed. The reported Microsoft spending projection was above $1 billion annually; a cut of more than a third would represent a substantial potential reduction if the estimate and change are accurate. That scale is not a reliable template for a smaller buyer. For a company spending $20,000 a month, the expense and disruption of switching could outweigh savings for some period, depending on its workloads and engineering capacity.
The practical budget question is therefore not simply which model charges less per token. Buyers need to compare cost per accepted result, including review, rework and operational overhead, and account for the cost of keeping another provider ready. The report does not provide enough information to calculate those costs for either company.
As an affiliate, we earn on qualifying purchases.
Alternatives Made the Moves Possible
Meta and Microsoft are not typical AI customers: each has substantial technology operations and a strategic reason to use alternatives. Meta develops its own models and coding tools. Microsoft owns GitHub Copilot and is a major backer of OpenAI. Redirecting employees to products a company owns, supports or already uses can reflect business strategy as well as supplier pricing.
That matters when interpreting the reported figures. A reduction in internal Claude use is not, by itself, evidence that an independent customer would make the same choice or that Claude has lost a head-to-head performance comparison. Nor does it show that the companies have stopped buying Anthropic technology for other purposes. The report describes a shift in some employee workflows while Microsoft’s reported customer-facing use remains in place.
For other organizations, moving between models can require new evaluations, prompt and tool adjustments, and integration work. Employees may need time to adapt. Coding agents can also depend on repository and editor integrations, while cached context and provider-specific pricing affect recurring costs. These are potential switching expenses identified in the source material; their size varies by system and workload.
“Meta reduced the number of employees using Claude Code from about 60,000 earlier this year to about 30,000.”
— The Information, in its Oct. 5 report
enterprise AI model management software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What the Report Does Not Establish
The report does not establish that Meta or Microsoft ended access to Claude, or that either company judged its quality to be inadequate. The specific savings achieved, the period over which they would accrue and the costs of changing tools are not detailed in the source material. Microsoft’s reported projection is not the same as verified realized spending or savings.
Some figures, including the account of monthly team budgets falling from around $100,000 to around $10,000, are attributed to a single report and may not represent every team. The source material also does not provide comparable evaluations of Claude, MetaCode, Muse Code, GitHub Copilot or OpenAI models on the companies’ internal tasks. The actual effect on productivity and total cost remains unclear.
As an affiliate, we earn on qualifying purchases.
What AI Buyers Can Track
For companies considering a similar move, the next step is to measure representative work on more than one model before shifting a large share of production use. Keeping a second provider active on a limited set of real tasks can expose integration and workflow costs earlier, while a company-owned evaluation set can show whether outputs meet its criteria.
Finance and engineering teams can track spending alongside accepted outputs, review time, rework and reliability. That comparison will help determine whether a cheaper model reduces total costs rather than only token charges. The report provides no announced date for further cuts or a broader change to either company’s Anthropic relationship, so the extent and durability of the reported shifts remain to be seen.
As an affiliate, we earn on qualifying purchases.
Key Questions
Are Meta and Microsoft ending their use of Claude?
The report describes reduced or redirected internal employee use, not an end to Claude access. Microsoft reportedly continues to use Anthropic models in customer-facing Copilot features.
Did the companies say Claude performs worse?
The source material does not report that either company cited inferior performance. The reported reasons include cost controls and available alternatives.
Does switching to another model automatically lower an AI budget?
No. A lower model or token charge can be offset by evaluation, engineering, integration, review and rework costs. The effect depends on the company’s workloads and the replacement’s performance on them.
What should a company measure before switching?
Compare model costs with the cost of usable, accepted results, including staff review time, rework, quality and any changes needed to prompts or integrations. Test on representative tasks before moving a large workload.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
