🔍 Read the full analysis: Switching From Claude: Costs, Trade-Offs, And What To Consider on ThorstenMeyerAI.com
Get office and shipping supplies 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 have reduced some employees’ use of Anthropic’s Claude tools, steering them toward products they own or already use. The reported shift concerns internal use, not a broad end to Claude access or customer-facing services, and the companies’ in-house alternatives make their switching costs unlike those of most businesses.
Meta and Microsoft have reportedly redirected some employees from Anthropic’s Claude tools toward alternatives, according to an Oct. 5 report by The Information. The reported changes concern the companies’ internal use, not a complete withdrawal of Claude from their products, and show how much a company’s ability to switch AI providers depends on having substitutes ready.
The Information reported that Meta cut the number of employees using Claude Code from about 60,000 earlier this year to about 30,000. The report said staff were steered toward Meta’s internal coding tools, MetaCode, which had more than 30,000 internal users, and Muse Code, with more than 6,000. The figures describe reported internal adoption; they do not establish that each former Claude user moved to one of those tools.
Microsoft had reportedly projected annual internal spending of more than $1 billion on Anthropic technology, including Claude Code, Claude models in Copilot and Claude Mythos. The report said Microsoft later cut that projection by more than a third and directed employees toward GitHub Copilot and OpenAI models. It also reported tighter token budgets; one account cited monthly team budgets falling from about $100,000 to about $10,000. That budget detail is based on a single report.
The reported reasons 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 says Microsoft continues to use Anthropic models for customer-facing Copilot features, while customer spending on Claude through Microsoft platforms is reported to be growing. The shift described is therefore a change in some internal use, not evidence that Claude access has ended.
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 Depends on Ready Alternatives
The report matters to organizations weighing AI costs because a lower model bill is only one part of the economics. Moving a production workflow can require teams to repeat evaluations, adapt prompts and tools, rebuild integrations and allow time for workers to learn another system. If a replacement performs less well on a company’s actual tasks, added review and rework can offset apparent savings.
Those costs are especially relevant for coding assistants, where value may depend on integration with an editor, code repository and team practices. A provider change can also alter caching behavior and pricing. For agent workloads that repeatedly use context, that may affect costs, but the source material gives no comparable measurements for Meta or Microsoft.
Meta and Microsoft have internal products and substantial engineering resources, which may make a move more feasible for them than for a company buying off-the-shelf tools. The report’s figures do not establish that switching will save money for other buyers. A business spending $20,000 a month, for example, cannot infer its own break-even point from large companies’ reported changes without calculating its integration, evaluation and productivity costs.
As an affiliate, we earn on qualifying purchases.
Internal Use Is Not a Customer Exit
Both companies have reasons to use alternatives they control or back. Meta develops its own models and coding tools; Microsoft owns GitHub Copilot and is a major backer of OpenAI. That competitive position is relevant when interpreting an internal purchasing decision: steering employees toward an in-house or affiliated product does not, by itself, show that a supplier’s technology failed.
The report, as summarized in the source material, distinguishes employee use from products offered to customers. Microsoft reportedly continues to use Anthropic models in customer-facing Copilot features, and Claude spending by customers through Microsoft platforms is said to be growing. These reported facts do not clarify how much customer usage or revenue is involved, but they caution against reading internal reductions as a company-wide break with Anthropic.
The broader operational issue is whether buyers can move work between providers without rebuilding everything. Running a second model on a limited share of real work, keeping business logic and prompts in a company-controlled layer, and maintaining representative evaluation tasks can make a later change easier to test. These are planning options, not steps the report says Meta or Microsoft took in full.
As an affiliate, we earn on qualifying purchases.
Reported Figures Leave Key Gaps
The available account does not provide the companies’ own detailed explanations or independently verified usage and spending data. It is unclear when Microsoft made the reported spending revision, how much it ultimately spent, and how the projected budget maps to actual usage. The reported employee counts also do not show how many people stopped using Claude altogether, used it less often or adopted multiple tools.
There is no task-by-task comparison of Claude and the replacement tools, nor evidence here that the shifts produced a measured productivity gain or loss. The effects of token prices, cache costs, review time and retraining are not quantified. The reported moves cannot establish that Claude is inferior, or that a similar switch would lower costs for a smaller buyer.
As an affiliate, we earn on qualifying purchases.
Measure Costs Before Switching
For companies considering a change, the next step is to compare models on representative work and accepted results, not token prices alone. Teams can test a limited set of real workflows, record quality and review needs, and include engineering time, integration work and any productivity dip in the comparison.
Further reporting or statements from Meta, Microsoft and Anthropic could clarify the timing, scale and stated reasons for the decisions. Until then, the confirmed picture is limited: the report describes internal shifts toward alternatives, while Microsoft reportedly retains Anthropic technology in customer-facing Copilot. Whether the changes deliver net savings, and what they mean for customer access, remains open.
As an affiliate, we earn on qualifying purchases.
Key Questions
Have Meta and Microsoft stopped using Claude?
No such complete stop is reported. The report describes reduced or redirected internal use. Microsoft is also reported to continue using Anthropic models for customer-facing Copilot features.
Why are employees reportedly being steered toward other tools?
The reported reasons include token costs, spending controls and available alternatives. The source material does not say either company attributed the changes to Claude performing worse.
Does the report show that switching away from Claude saves money?
No. It reports a reduced Microsoft spending projection, but does not provide a full accounting of migration costs, productivity effects or realized savings. Results for another company may differ.
What costs should a company include before changing models?
Consider evaluation work, prompt and integration changes, employee learning time, cache economics, review and rework. Token spend alone may not reflect the full cost of completing work to an acceptable standard.
What information is still missing?
The public account does not establish final spending, detailed usage changes, comparative performance on company tasks or the net effect on productivity. Further company statements or reporting may clarify those points.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
