Businesses Reevaluate AI Spending as Costs Continue to Climb

NEW YORK: (Web Desk) – Artificial intelligence is becoming increasingly costly, prompting many businesses to reconsider how heavily they rely on the rapidly evolving technology.

In the early days of the AI boom following the launch of ChatGPT, major technology firms offered services at relatively low prices, supported largely by investor funding. Industry experts describe this phase as one of “subsidized intelligence,” where companies focused on attracting users rather than generating immediate profits.

However, that strategy is beginning to change. As leading AI developers such as OpenAI and Anthropic pursue long-term profitability and potential public market listings, pressure is mounting to turn AI products into sustainable businesses.

A key factor behind rising costs is the growing use of AI agents. Unlike traditional chatbots that simply respond to questions, AI agents can perform complex tasks such as scheduling appointments, writing software, managing files and handling multi-step workflows. These activities require significantly more computing power and consume far more tokens, the units used by AI companies to measure and charge for usage.

At the same time, demand for advanced computer chips and data-center infrastructure continues to outpace supply, increasing operational costs across the industry.

Technology consultant Mark Barton noted that AI-related expenses, particularly for software development and coding applications, have risen sharply as businesses expand their usage.

Some organizations have embraced AI so aggressively that they have encountered a phenomenon known as “tokenmaxxing,” where excessive use of AI tools drives costs higher than expected.

According to analyst Jack Gold, some companies have discovered that AI token expenses can surpass the cost of employing staff members within a relatively short period due to excessive consumption.

Even major technology companies are taking a more cautious approach. Reports suggest that Meta has begun encouraging employees to use AI more strategically rather than simply maximizing usage as a measure of productivity.

Questions about the return on investment from AI spending are also growing. Uber’s chief operating officer recently suggested that significant investments in AI have yet to deliver a noticeable boost in productivity.

Trump Claims Iran Agreement Close Despite Unresolved Key Issues

In response, businesses are exploring ways to control costs. Many are turning to free and open-source AI models that provide acceptable performance for routine tasks without the premium price tag of advanced proprietary systems.

Others are adopting smaller, industry-specific AI models tailored for sectors such as finance, healthcare and real estate. Companies are also breaking large AI workloads into smaller tasks and assigning each step to the most cost-effective model available.

Experts say the savings can be substantial. While advanced flagship models may cost around $15 per million tokens, smaller models can reduce that figure to just a few cents for the same volume of usage.

The shift reflects a broader trend toward treating AI as a commodity, where selecting the most efficient and affordable model becomes more important than relying solely on the most powerful systems.

Despite these changes, industry observers believe premium AI models will continue to attract businesses that require cutting-edge capabilities, ensuring strong demand for advanced services even as the market becomes more cost-conscious.