AI was supposed to save money, so why are costs rising? Companies worldwide are grappling with these challenges.

Why AI is Getting Expensive: AI was viewed as a cost-cutting technology, yet it has driven up expenses. Consequently, companies are now exploring cheaper alternatives.

Why AI is Getting Expensive: AI has recently dominated the landscape. Companies are replacing employees with AI to handle tasks, regarding it as the ultimate productivity tool. While companies are heavily prioritizing AI to cut operating costs and boost efficiency, are expenses actually decreasing? This question arises because AI spending continues to rise, and the promise of cost savings does not appear to be fully materializing.

Using AI is becoming expensive.

AI is not a technology where a one-time investment suffices. Companies incur massive monthly expenses on subscriptions, enterprise licenses, usage-based pricing, and employee training. Recently, reports emerged that a company spent approximately ₹4,200 crore on Anthropic’s Claude. Similarly, Amazon has advised its employees to use AI judiciously to keep bills in check. Uber’s COO has stated that AI usage is prohibitively expensive and difficult to justify; Uber exhausted its entire annual AI budget in just four months.

Why are costs rising?

Unlike traditional software, which can be purchased via annual licenses, AI operates differently. Many generative AI services function on subscription models or token consumption. With AI now being used for everything from writing to data analysis, costs are spiraling. This is why companies are now seeking ways to cut these expenses. Many companies are now shifting from premium AI models to smaller, more affordable ones.

AI's limitations are also coming to light.

There is no doubt that AI is a powerful tool, but as its usage grows, its limitations are becoming apparent. AI cannot perform every task that requires human input. A recent example involves Ford Motor Company; following AI integration, the company laid off 350 engineers over the past three years, only to find that AI could not match their performance. AI lacks both the skills and the experience. Consequently, the company is now rehiring those engineers.