Is AI On The Way Down?

New Delhi, Oct. 25 -- "Stock market bubbles

don't grow out of thin air."

- George Soros

The artificial intelligence revolution was supposed to rewrite the rules of business. It may yet do so. But a more uncomfortable question is beginning to emerge: what if AI changes the world without making all those betting on it rich?

Two recent bankruptcy filings offer a sobering warning. Getty Images, the visual-content giant, entered Chapter 11 proceedings in the US on October 21 to restructure its debt and move towards private ownership. Character.AI, a five-year-old chatbot platform, initiated its own Chapter 11 process earlier this month, seeking a court-supervised sale amid legal, regulatory and operational challenges.

Neither filing means the businesses will disappear. But together, they expose two vulnerabilities in the AI economy. Getty represents the established business confronting technological disruption. Character.AI represents the new generation of AI-powered businesses struggling to make their popularity commercially sustainable. One is threatened by the technology transforming its market; the other must find a way to make that technology pay.

When Disruption Bites

Getty built its business by licensing photographs and visual content to publishers and advertisers. Generative AI can now produce images from text prompts, potentially reducing demand for some traditional stock imagery. The threat is not that every AI-generated image can replace a professional photograph for news and documentary purposes. It is that customers now have alternatives.

Getty's financial difficulties also reflect substantial debt and other business pressures. Its restructuring cannot be attributed to AI alone. Nevertheless, the company illustrates how technological disruption can weaken the economics of an established business even before its replacement becomes universal.

Character.AI presents the opposite problem. Its chatbots attracted users by offering personalised conversations and simulated relationships. Yet popularity is not the same as profitability. The company must contend with the expense of running AI services, alongside litigation and scrutiny of user safety and privacy. The lesson is uncomfortable for both incumbents and newcomers. AI can undermine an existing revenue model, while making it difficult for a new business to build one that generates adequate returns.

Cost of Intelligence

The financial strain extends to the companies developing the most advanced AI models. Training and operating these systems require expensive infrastructure, specialised chips, energy and talent. More sophisticated models and growing usage can increase revenues, but they can also increase costs.

Anthropic, one of the leading players in the sector, reportedly recorded an operating loss exceeding $8 billion against revenue of $4.6 billion in 2025. Such figures are not unusual for an expanding tech firm investing heavily in future growth. Nor do they establish that the business model is doomed. But they demonstrate the enormous financial commitment required to compete at the frontier of AI.

OpenAI offers another reason to examine the numbers. Its reported annualised revenue was nearing US $50 billion at the end of September, below the $70-billion figure previously reported. Part of the discrepancy arose from differences in how OpenAI and Anthropic calculate revenue generated through cloud partners. OpenAI's figure still indicates substantial growth, but the episode highlights the danger of treating headline numbers as straightforward measures of comparable performance.

Growth Is Not Profit

This is where AI faces a defining test. Enormous investment is flowing into data centres, chips, models and applications on the expectation that AI will transform productivity and create new markets. That expectation is not fanciful. AI is already useful in coding, research, customer service, content creation and numerous business processes.

But a transformative technology does not guarantee attractive returns for every company producing or deploying it. Competition can push down prices. Computing costs can absorb revenue gains. Customers may experiment without paying enough to support the services they use. The companies supplying the infrastructure may themselves be vulnerable if demand fails to justify the investment.

The internet boom offers a parallel. The dotcom bubble burst around 2000, destroying wealth. Yet the internet did not disappear. It became the foundation of some of the world's most valuable businesses. The mistake was not believing in the internet; it was assuming that every business built around it would succeed and that every valuation was justified. AI could follow a similar path.

Reality Check

So, is AI on the way down? The evidence does not yet justify that conclusion. Two bankruptcy proceedings and concerns over revenues do not mean an entire industry is in terminal trouble. But they do challenge the notion that technological progress will always translate into financial success.

This distinction matters. AI investment influences global markets, spending and corporate strategies. A correction could affect investors and businesses beyond the companies developing AI, even as adoption continues. The question is not whether AI is revolutionary. It is whether the prices being paid for its promise adequately reflect the costs, risks and time required to realise that promise.

A technology can be extraordinary. Its business model can be flawed. Its valuation can be excessive. All three can be true at once. The AI revolution may survive any market correction. The real test is how many of today's celebrated businesses can survive the transition from promise to profit.

Views expressed are personal. The writer is a journalist & communications specialist

Published by HT Digital Content Services with permission from Millennium Post.