OpenAI slows revenue and new user growth amid expensive AI infrastructure costs

OpenAI is entering its most challenging phase since ChatGPT's explosive success: the market has already proven its interest in generative AI, but now the company must demonstrate that it can convert this interest into sustainable revenue faster than computing costs grow. Anxiety is intensifying within the startup because actual revenue growth and the influx of new users are falling short of target benchmarks, and part of the audience is moving to competitors.

The problem is especially acute against the backdrop of OpenAI's capital-intensive model. The company is simultaneously developing new models, expanding its product line, and forced to maintain enormous computational capacity for training and running services. This requires not only research funding but also constant investments in data centers, chips, network infrastructure, and leasing of existing cloud resources. The more popular AI products become, the greater the load and the stronger the dependence on external financing.

ChatGPT gave OpenAI a powerful start and made the brand almost synonymous with consumer AI. But after the initial wave of hype, the market became much more crowded. Users are no longer confined to a single service: they have a choice between major platforms, niche AI tools, embedded assistants in office products, and open models that can be run locally or through cheaper cloud interfaces.

In such an environment, growth can no longer be sustained by novelty alone—the company must retain its audience and prove practical value in everyday use cases.

The slowdown in the rate of new user acquisition is particularly painful for OpenAI because the company's expenses are inherently difficult to scale down. If demand grows slower than expected, this does not automatically reduce costs: infrastructure, cloud provider contracts, and development teams still need to be paid for.

As a result, the gap between the ambitions embedded in growth plans and actual monetization becomes a strategic problem, not merely a temporary fluctuation in metrics. The exodus of some users to competitors compounds this picture.

In the generative AI segment, switching between services happens more easily than in many mature digital markets. If comparable response quality, a more convenient interface, lower pricing, or fewer restrictions are available in another product, a significant portion of the audience is willing to experiment.

This means that for OpenAI, the competition is no longer only about model quality per se, but also about product speed, stability, transparent pricing, integrations with work tools, and clear ROI scenarios for business.

This is also an important signal for investors and partners. Until now, the willingness to invest in OpenAI has been largely sustained by the belief that research leadership automatically converts into market dominance. But as the industry matures, more pragmatic questions come to the fore: what is the cost of acquiring a user, how long do they remain active, how predictable is revenue growth, and can infrastructure be scaled without endlessly accelerating expenses.

The AI market is increasingly evaluating not only technological breakthroughs but also operational discipline.

A separate challenge relates to the enterprise segment. For business, impressive demos are not enough; what matters is security, cost control, legal predictability, integration quality, and service stability under load.

If OpenAI wants to compensate for the slowdown in the consumer segment, it needs to strengthen exactly those offerings that turn the model into a working tool within companies rather than a one-time capability demonstration. This typically delivers more stable revenue but requires lengthy sales cycles and serious support.

The situation around OpenAI shows that the generative AI market is entering a more competitive phase. The winner will not be the one who first amazed the world, but the one who can simultaneously retain its audience, sell clear value, and control infrastructure costs. For OpenAI, this is a moment of maturity testing: the brand and technological leadership still provide a strong position, but going forward it must be confirmed by business economics, not merely by the loudness of new model launches.