Anthropic, OpenAI customers overbilled by $1.7 million due to billing errors

Anthropic, OpenAI customers overbilled by $1.7 million due to billing errors

Companies relying on AI models and services are paying an estimated $1.7 million in overcharges due to billing errors.

Advertisement
    Share:
Since AI services are billed based on usage, and such cost fluctuations could have a serious impact on how companies plan to rely on AI.Since AI services are billed based on usage, and such cost fluctuations could have a serious impact on how companies plan to rely on AI.
Business Today Desk
  • Jun 29, 2026,
  • Updated Jun 29, 2026 11:19 AM IST

Recently, several companies have deployed AI tools, agents, and models from OpenAI and Anthropic across operations to streamline their production processes. However, many companies using these services have been billed incorrectly. According to The Information, companies relying on AI models and services are paying an estimated $1.7 million in overcharges.

Advertisement

A startup called Vaudit, which specialises in auditing AI bills, said to review $34 million worth of AI invoices from 60 companies between March and June 2026. During the audit, the company found a major billing overcharge. The majority of the billing issues were related to Anthropic's Claude Code, although it also found invoices involving OpenAI services.

Must read: US eases block on Anthropic’s Mythos 5, allowing limited release to key cyber defenders 

The startup company said that it could not independently verify AI invoices because pricing depends on several factors, including the type of AI model, token usage, API calls, and others. However,  it says several major companies, including Panasonic, HP, and Honda, are among its customers.

What may have caused billing errors?

Advertisement

Laura Bratton from The Information explained the reason behind the overbilling, saying that “Maybe there was an AI agent that continued to retry a task that it was completing incorrectly, and a customer didn't realise that the AI agent kept trying over and over again, and they were getting charged for it.”

However, Anthropic and OpenAI both reportedly denied that billing errors are occurring, saying they have not seen evidence supporting those claims. The report comes at a time when companies are already spending massive amounts to use generative AI models. Since AI services are billed based on usage, and such cost fluctuations could have a serious impact on how companies plan to rely on AI.

Must read: 'Why argue with someone when you don't have…': Anthropic CEO Dario Amodei on why he left OpenAI

Advertisement

Vaudit further revealed that after customers raised complaints about overcharged AI bills, AI providers, including Amazon, Google, Microsoft, Anthropic, and OpenAI, refunded around 80% of the disputed charges. However, despite the issues, companies have continued to use premium AI models. On the other hand, many are also reported to look for cheaper alternatives. However, it also raises the need for transparent and accurate billing amid growing AI compute and prices.

For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine

Recently, several companies have deployed AI tools, agents, and models from OpenAI and Anthropic across operations to streamline their production processes. However, many companies using these services have been billed incorrectly. According to The Information, companies relying on AI models and services are paying an estimated $1.7 million in overcharges.

Advertisement

A startup called Vaudit, which specialises in auditing AI bills, said to review $34 million worth of AI invoices from 60 companies between March and June 2026. During the audit, the company found a major billing overcharge. The majority of the billing issues were related to Anthropic's Claude Code, although it also found invoices involving OpenAI services.

Must read: US eases block on Anthropic’s Mythos 5, allowing limited release to key cyber defenders 

The startup company said that it could not independently verify AI invoices because pricing depends on several factors, including the type of AI model, token usage, API calls, and others. However,  it says several major companies, including Panasonic, HP, and Honda, are among its customers.

What may have caused billing errors?

Advertisement

Laura Bratton from The Information explained the reason behind the overbilling, saying that “Maybe there was an AI agent that continued to retry a task that it was completing incorrectly, and a customer didn't realise that the AI agent kept trying over and over again, and they were getting charged for it.”

However, Anthropic and OpenAI both reportedly denied that billing errors are occurring, saying they have not seen evidence supporting those claims. The report comes at a time when companies are already spending massive amounts to use generative AI models. Since AI services are billed based on usage, and such cost fluctuations could have a serious impact on how companies plan to rely on AI.

Must read: 'Why argue with someone when you don't have…': Anthropic CEO Dario Amodei on why he left OpenAI

Advertisement

Vaudit further revealed that after customers raised complaints about overcharged AI bills, AI providers, including Amazon, Google, Microsoft, Anthropic, and OpenAI, refunded around 80% of the disputed charges. However, despite the issues, companies have continued to use premium AI models. On the other hand, many are also reported to look for cheaper alternatives. However, it also raises the need for transparent and accurate billing amid growing AI compute and prices.

For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine

Read more!
Advertisement