🔍 Read the full analysis: Claude To Another AI Model: How To Think Through Switching Costs on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
The Information reported on Oct. 5 that Meta and Microsoft have reduced some employees’ use of Anthropic’s Claude tools while directing them to internal or other products. The reported changes concern internal use, not an end to Claude access or a broad customer departure; they highlight how cost, available substitutes and the work of switching shape enterprise AI choices.
Meta and Microsoft are steering some employees away from Anthropic’s Claude tools and toward alternatives, according to an Oct. 5 report by The Information. The reported moves concern the companies’ internal use, not a decision to end access to Claude, and show how cost and ready-made substitutes can influence enterprise AI spending.
The Information reported that Meta reduced the number of employees using Claude Code from about 60,000 earlier this year to about 30,000. It said Meta has been directing staff toward its own coding tools: MetaCode, which has more than 30,000 internal users, and Muse Code, with more than 6,000.
Microsoft had 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. It has since cut that projection by more than a third and is steering employees toward GitHub Copilot and OpenAI models. The report also says Microsoft continues to spend on Anthropic models for customer-facing Copilot features, while customer spending on Claude through Microsoft platforms is growing.
The reported reasons for the internal shifts include rising token costs, tighter spending controls and the availability of tools the companies own or support. The source material does not report either company saying Claude performed worse. It also describes stricter Microsoft token budgets; one account cited monthly team budgets dropping from about $100,000 to about $10,000, a detail attributed to a single report.
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 Existing Alternatives Matter
The changes are a reminder that a model’s listed price is only one part of what an organization pays. Moving work also carries engineering, evaluation and productivity costs. Teams may have to retest workflows, adapt prompts and tools, and allow employees time to learn a different system. A replacement that performs less well on a company’s particular tasks could add review and rework, even if its token price is lower.
Meta and Microsoft appear to have had substitutes ready to use. That makes their choices different from those facing many smaller buyers, which may lack internal coding products or the staff to build and maintain them. The report does not establish that every customer can save money by switching, or that one model is better. It shows that the ability to route work elsewhere can matter when costs or commercial priorities change.
For buyers, the practical implication is to measure total cost per accepted result, not just tokens or subscriptions. Maintaining a second option, tracking representative task performance and keeping business logic separate from a particular provider can make a future change less disruptive. Those steps have costs too; the case for them depends on a company’s scale, risk and expected savings.
As an affiliate, we earn on qualifying purchases.
Internal Buyers and Rival Products
Both companies have reasons to develop or promote alternatives. Meta builds its own models and coding tools. Microsoft owns GitHub Copilot and is a major backer of OpenAI. When a company also sells or develops competing products, directing its own staff toward them can reflect commercial alignment and vertical integration, as well as spending considerations.
The reported figures describe employee adoption and an internal spending projection, not a complete accounting of each company’s Anthropic relationship. Microsoft’s continued use of Claude for some customer-facing features illustrates why internal tool choices should not automatically be read as a company-wide supplier exit. The source material also describes no announcement ending either firm’s access to Claude.
Switching can involve more than changing an API setting. Evaluations need to be repeated, prompts and tool definitions may need adjustment, and integrations with editors and repositories may not transfer cleanly. Teams can also lose cached context or face different cache pricing. These are potential costs outlined in the source material, not quantified outcomes reported for Meta or Microsoft.
As an affiliate, we earn on qualifying purchases.
What the Report Does Not Establish
The source material does not provide company statements confirming the reported figures or explaining the decisions in the companies’ own words. It is also not clear exactly when each usage change took effect, how the employee counts were measured, or how the spending projection was calculated.
The report, as presented here, does not quantify savings against the full cost of switching, including engineering time, evaluation work and any change in output quality. It does not establish that Claude underperformed on either company’s tasks, nor that the reported moves extend to customer products or external customers. The Microsoft team-budget figures are described as coming from a single account, so their scope and representativeness remain unclear.
As an affiliate, we earn on qualifying purchases.
Track Adoption and Total Costs
The next useful indicators are whether Meta and Microsoft continue shifting internal work, how their internal alternatives perform on real tasks, and whether reported spending reductions persist after migration costs are counted. Further company comments or reporting could clarify the timeline, the role of token budgets and the scope of Microsoft’s continuing Anthropic use.
For other organizations, the immediate next step is not necessarily to switch providers. It is to establish a baseline: test representative tasks on more than one model, record the human review and rework required, and keep track of integration costs. That evidence can show whether a move saves money in practice—or whether the cost of changing tools outweighs the expected benefit.
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 access. It says Microsoft continues to use Anthropic models for some customer-facing Copilot features.
Why are the companies reportedly shifting employees?
The reported factors include token costs, spending controls and available alternatives. The source material does not say either company declared Claude inferior in quality.
Does this show Claude is worse than the alternatives?
No. The reported decisions do not establish comparative performance. Evaluating that would require task-specific results and information about review, rework and overall cost.
What makes switching AI tools expensive?
Potential costs include retesting workflows, adapting prompts and integrations, retraining users, and changes in review time or output quality. The source material does not quantify those costs for Meta or Microsoft.
What should other companies learn from the report?
Organizations can compare models on representative tasks and track the full cost per accepted result. Keeping a tested alternative available may reduce disruption, but whether that investment pays off depends on the company’s needs and scale.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
