Claude reportedly bills a client $500M for one month of AI use
As more companies begin to question the cost-benefit ratio of using AI, this one has a much bigger problem to deal with. A report claims that the company may have been charged $500 million for one month of Anthropic’s Claude.
Listen to this article
As more companies begin to question the cost-benefit ratio of using AI, this one has a much bigger problem to deal with. A report claims that the company may have been charged $500 million for one month of Anthropic’s Claude.
- A report claims a company may have received a $500M bill for one month of Claude AI use.
- The spike was reportedly caused by unlimited employee access with no usage caps.
- Firms are increasingly discovering AI is far more expensive than expected at scale.
- Microsoft and Uber have already cut or exhausted AI budgets due to soaring costs.
Key Takeaways by nexos.ai, reviewed by Cybernews staff.
While the report hasn't been officially confirmed, Axios cites an undisclosed AI consultant, saying that one of their clients recently failed to set usage limits on Anthropic's Claude licenses for employees, resulting in a half-a-billion-dollar monthly bill.
The report didn't specify what happened next or whether the bill has been paid or will be paid at all. In either case, both individual and enterprise users of AI platforms are now beginning to see more clearly the technology's current limitations and learning to use it more efficiently.
Have thoughts about this topic? Others do, too. Join them in the discussion.
Enterprise users are starting to notice that their employees default to automating tasks they dislike instead of those that are really valuable to the company, with some even using AI to check the weather, increasing costs for businesses. What's more, the growing consensus seems to be that AI, at the enterprise level, is currently most useful for coding. Also, while companies are learning to work with AI, layoffs are now cited as the main lever CEOs can use to offset rising AI costs.
And this is all happening at a time when sentiment across the general population, at least in the US, toward AI expansion is growing more negative, as people worry about the rapid expansion of this technology and its impact on their jobs.
In either case, this steep learning curve for enterprise AI users has already forced some tech giants to review their strategy.
For example, as reported by Cybernews, after inviting thousands of developers to use Anthropic's Claude Code in December, Microsoft exhausted its annual AI budget due to token-based billing and is now terminating licenses by June 30th.
Developers are now being steered toward Copilot CLI, an AI-powered coding assistant developed by Microsoft-owned GitHub.
Another company that ran out of money for AI before the year is even halfway through is Uber. The company spent its entire 2026 AI budget in just 4 months.
FAQ
Does Claude Code have a free tier?
Claude Code does not have a free tier. It requires a paid subscription to use, typically through Anthropic's Claude Pro or API access plans.
Does Claude Code run locally?
Claude Code (the CLI tool) can be configured to run with local open-source models on your machine, though by default it uses Anthropic's cloud API. So yes, local execution is possible when you set it up with compatible local models.
Is Claude Code open source?
No, Claude Code is not open source. It uses a proprietary license despite being hosted on GitHub. However, there was an accidental source code leak in April 2026 when Anthropic shipped a sourcemap in their npm package, which some developers have used to create unofficial open-source rebuilds.
Unlock more exclusive Cybernews content on YouTube.