AI Spending Per Employee Hits $7,400 at Top Companies

Companies are rapidly increasing their spending on Artificial Intelligence as AI tools become a bigger part of everyday work. A new report suggests that the world’s most aggressive corporate AI adopters are spending dramatically more per employee than the average Business.

According to Ramp’s AI Index, the top 1% of companies in the US spent a median $7,400 (roughly Rs 7 lakh) per employee on AI last month.

That represents a substantial increase from the beginning of the year. In January, companies in the top 1% were spending a median $2,590 (around Rs 2.47 lakh) per employee on AI.

The figures highlight a growing divide between companies that are Investing heavily in AI and those that are still using the Technology on a relatively limited scale.

AI spending has surged in recent months

The increase from $2,590 per employee in January to $7,400 last month represents a significant jump in AI-related spending among the most aggressive adopters.

The trend comes as businesses increasingly provide employees with access to AI assistants, coding tools, research platforms and other generative AI services.

However, these tools can become expensive when used extensively across large organisations.

AI providers such as Anthropic and OpenAI have increasingly adopted usage-based pricing models, including charges based on the number of tokens processed.

In simple terms, the more work an employee performs through an AI system, the more tokens may be consumed, potentially increasing the company’s bill.

For businesses deploying AI at scale, that means costs can rise rapidly as employees use the tools for increasingly complex and frequent tasks.

Top 10% spend far less than the top 1%

The Ramp data shows that AI spending varies significantly between companies.

While the top 1% of companies were spending a median $7,400 per employee, the top 10% were spending approximately $650 (around Rs 62,000) per employee.

The overall median was considerably lower at just $11.95 (roughly Rs 1,140) per employee.

The figures illustrate the enormous difference in AI adoption and spending across businesses.

At one end are companies making AI a major part of their operations and spending thousands of dollars per employee. At the other are businesses where AI-related spending remains relatively small.

According to the report, this gap between heavy AI investors and companies using AI more lightly is widening.

Why are AI costs increasing?

One reason is the growing use of AI tools for more demanding tasks.

Employees are no longer using AI only for simple text generation or basic queries. Companies are increasingly deploying AI for coding, research, analysis and other workflows that can require significantly more computing resources.

Usage-based pricing means that greater AI adoption can directly translate into higher costs.

For companies with thousands of employees, even relatively small increases in spending per employee can therefore become significant expenses.

This creates a balancing act for businesses: they want employees to use AI enough to improve productivity, but they also need to ensure that the cost of AI does not grow faster than its business benefits.

Some companies are already limiting AI usage

The rapid increase in AI-related costs has also prompted some major companies to introduce restrictions.

According to the report, Uber, Amazon and Walmart have introduced limits on token usage for employees in an effort to control spending.

Such restrictions reflect a growing realisation among businesses that giving employees unlimited access to AI tools can result in unpredictable costs.

Instead, companies may increasingly monitor usage and establish limits based on the type of work being performed.

For businesses experimenting with AI at scale, managing usage could become just as important as choosing which AI model or platform to use.

Sam Altman has highlighted the cost challenge

The rapid rise in AI spending has also become a talking point among technology executives.

OpenAI CEO Sam Altman previously highlighted the issue while describing the growing pressure companies face as AI usage expands.

He referred to a hypothetical situation in which a company had already spent its entire annual budget during the first quarter and was asking how it could make AI more efficient.

The comment reflects the tension between rapidly increasing AI adoption and the costs associated with running increasingly powerful models.

Compute costs can exceed employee costs

The cost of AI Infrastructure has also been highlighted by Nvidia executive Bryan Catanzaro.

In April, Catanzaro said that for his team, the cost of computing was far higher than the cost of the employees themselves.

That observation underlines how unusual the economics of AI can be compared with traditional software.

Historically, hiring additional employees has generally been one of the largest expenses for many technology companies. With AI, computing resources can become a major cost centre in their own right.

As companies use more sophisticated models and process larger amounts of information, that computing bill can grow rapidly.

Anthropic leads business AI adoption

The Ramp data also provides insight into which AI providers businesses in the US are using.

According to the report, Anthropic continued to lead AI adoption among US businesses, with a 43.5% share of AI-using companies.

OpenAI followed with 39.7%.

Elon Musk’s xAI accounted for around 4%, according to the report.

These figures show that the business AI market remains concentrated around a small number of major providers, even as more alternatives emerge.

Open-source and Chinese models are gaining ground

The report also points to increasing use of model-serving platforms that provide access to open-source models and some Chinese AI models.

Their share rose to 6.1% of AI-using businesses in July, according to Ramp.

The report said this trend has not yet meaningfully reduced spending on OpenAI and Anthropic.

However, it could indicate a change in how heavy AI users manage their costs.

Instead of relying entirely on one commercial AI provider, companies may increasingly combine premium models with cheaper open-source alternatives depending on the task.

For example, businesses could use more expensive models for complex reasoning or high-value workflows while turning to lower-cost models for simpler tasks.

The AI spending gap could get even wider

The most striking takeaway from the Ramp data is not simply that companies are spending more on AI.

It is that AI spending is becoming increasingly uneven.

The top 1% of companies spend a median $7,400 per employee, compared with $650 for the top 10% and just $11.95 across the overall median.

That suggests some companies are moving aggressively toward AI-heavy operations while others remain at an early stage of adoption.

The difference could reflect factors such as company size, industry, employee roles and how deeply AI has been integrated into business processes. The supplied report does not provide a detailed breakdown of those factors.

What higher AI spending means for businesses

For companies, the question is increasingly shifting from whether to use AI to how much AI to use and how to control the costs.

Heavy spending could make sense if AI tools generate enough productivity gains or new revenue to justify the investment.

But simply increasing the number of AI subscriptions or allowing unlimited model usage does not automatically guarantee better results.

Businesses therefore face a new challenge: measuring whether their AI expenditure is producing meaningful returns.

The introduction of token limits by companies such as Uber, Amazon and Walmart suggests that cost management is already becoming part of corporate AI strategy.

AI’s next phase could be about efficiency

The rapid increase in spending among the biggest AI adopters suggests that the initial phase of corporate AI adoption was focused heavily on access.

Companies wanted employees to experiment with AI and discover how the technology could improve workflows.

The next phase could focus much more on efficiency, usage monitoring and return on investment.

As AI models become more capable, their use cases will likely expand. But so too can the costs associated with processing increasingly complex workloads.

The Ramp figures show just how different those costs can be from one company to another.

For the most aggressive adopters, AI is already becoming a major line item, with spending reaching thousands of dollars for every employee. For the typical company, however, AI expenditure remains much smaller.

That widening gap could become one of the defining trends in corporate AI adoption as businesses determine whether bigger AI budgets actually translate into bigger business results.