🔍 Read the full analysis: What Meta And Microsoft’s Claude Pullback Reveals About Switching Costs on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
The Information reported on Oct. 5 that Meta and Microsoft have reduced some employees’ use of Anthropic’s Claude tools while steering them toward alternatives they own or back. The reported changes concern internal use, not a withdrawal of Claude from customer-facing products, and illustrate how switching models is easier for companies with working substitutes than for most buyers.
Meta and Microsoft are steering some of their own engineers away from Anthropic’s Claude tools and toward alternatives, according to an Oct. 5 report by The Information. The reported moves concern internal use, not a broad end to Claude access, and show how a company’s ability to switch AI providers depends on whether it already has substitutes deployed.
Meta reportedly reduced Claude Code use from about 60,000 employees earlier this year to about 30,000. The report says the company has been directing staff to its own coding tools: MetaCode, which has more than 30,000 internal users, and Muse Code, which has more than 6,000.
Microsoft reportedly cut its internal spending projection on Anthropic technology by more than a third from a level above $1 billion a year. The spending included Claude Code, Claude models in Copilot and Claude Mythos. The report says Microsoft has redirected employees toward GitHub Copilot and OpenAI models. It also describes tighter token budgets; one account cited by the source material put some monthly team budgets at about $10,000, down from about $100,000. That detail is based on a single report.
Cost controls and competing products are the reported reasons for the internal changes. Neither company is reported to have said Claude performed worse. Microsoft is also reported to continue spending on Anthropic models for customer-facing Copilot features, while customer spending on Claude through Microsoft platforms is said to be growing. The reporting does not indicate that either company has ended access to Claude.
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 matter less as a verdict on Claude than as a case study in who can change providers without disrupting work. Meta and Microsoft have alternatives already in use: Meta’s coding tools, and Microsoft’s GitHub Copilot and access to OpenAI models. Their scale and existing engineering investment may make it easier to shift workloads when costs or internal priorities change.
For other companies, a model’s listed price is only part of the cost. Switching can require retesting workflows, adapting prompts and integrations, and retraining users. A new model may also affect output quality, review time and rework. Those costs are specific to each organization; the report does not establish what either company saved after accounting for them.
The practical implication is not that businesses should leave Claude or choose a particular rival. It is that dependence on a single model can limit a buyer’s options. Maintaining a second provider on real tasks and tracking quality against representative evaluations can make future changes more manageable, though it still takes engineering and operating effort.
As an affiliate, we earn on qualifying purchases.
Internal Use Versus Customer Products
The reported numbers refer to employees’ internal use and company spending projections. They should not be read as evidence that Meta or Microsoft customers have lost access to Claude. The source material says Microsoft continues to use Anthropic models in some customer-facing Copilot features and that Claude spending through Microsoft platforms is growing.
Both companies also have a commercial stake in alternatives. Meta develops its own models and coding tools; Microsoft owns GitHub Copilot and is a major backer of OpenAI. Directing employees to a company’s own product or an affiliated provider can reflect cost management and business strategy as well as model selection. The available reporting does not isolate how much each factor contributed.
The figures are reported observations, not a direct comparison of model performance. Meta’s user totals describe reported internal adoption at different points in time, while Microsoft’s figure is a spending projection and reported revision. They measure different things and do not, by themselves, show that one model is more capable or cost-effective for every task.
As an affiliate, we earn on qualifying purchases.
Savings and Model Performance
The reported account does not establish net savings after engineering, evaluation, retraining and any productivity effects are included. It also does not provide a like-for-like comparison of Claude with the alternatives on Meta’s or Microsoft’s actual workloads. Neither company is reported to have cited poorer Claude performance as the reason for the shift.
It remains unclear how quickly the user and spending changes took effect, which teams or tasks were affected, and how much internal Claude use continues. The reported team-budget figures come from a single account, and the underlying measurement periods and methods are not detailed in the supplied reporting. The companies’ detailed explanations and complete figures are not provided here.
As an affiliate, we earn on qualifying purchases.
Watch for Updated Usage Data
The next useful evidence would be updated figures from Meta and Microsoft on employee adoption, actual Anthropic spending and the internal alternatives now handling the work. More detail about the affected teams and tasks would help distinguish broad replacement from targeted cost controls.
For other buyers, the relevant test is their own: compare models on representative tasks, include review and rework in cost calculations, and check how much effort a change would require. Until the companies publish more detail, the reported pullback supports a limited conclusion: having a working alternative can make switching more feasible, but the economics and performance depend on the workload.
As an affiliate, we earn on qualifying purchases.
Key Questions
Have Meta and Microsoft stopped using Claude?
The report does not say either company has ended Claude use. It describes reduced or revised internal use and continued Microsoft use of Anthropic models for some customer-facing Copilot features.
Why are the companies steering employees to other tools?
The Information reported cost pressures, tighter spending controls and a preference for tools the companies own or are invested in. It did not report that either company said Claude performed worse.
Does the report show that Claude is less capable?
No. The reported usage and budget changes do not provide a direct model-performance comparison. The reasons attributed in the report center on cost and available alternatives.
Why might switching be harder for a smaller company?
A smaller buyer may not already have another model integrated into its workflows. Changing providers can mean retesting tasks, modifying prompts and software, and helping staff adjust; the time and cost vary by company and use case.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
