Three years after the debut of the hugely popular AI chatbot ChatGPT, OpenAI, a startup valued at $500 billion, is facing substantial hurdles in maintaining its lead within the AI market, as reported by The Financial Times. Previously holding a significant advantage over rivals, OpenAI is now experiencing considerable pressure from both Google and Anthropic, who have recently solidified their market positions.
Image Credit: Growtika/unsplash.com
Image Credit: Growtika/unsplash.com
Last week, Google unveiled its large language model, Gemini 3, which purportedly surpassed OpenAI’s GPT-5 across several crucial metrics, outperforming other competitor AI models as well. It achieved levels of model training efficacy that OpenAI has struggled to replicate in recent months.
Even prior to this, OpenAI CEO Sam Altman had alerted staff via an internal memo that the company “will need to maintain focus amid short-term competitive pressure… The market situation is expected to be challenging.”
A year ago, many observers doubted Google’s capacity to bridge the massive AI gap separating it from OpenAI. However, a turning point arrived early this year, following a series of updates showcased at the Google I/O 2025 developer conference in May, and driven significantly by the well-received photo-editing tool, Nano Banana AI. The monthly user base for the Gemini mobile application surged to 650 million, up from approximately 400 million in May. This progress immediately fueled a sharp climb in Alphabet’s stock, pushing the holding company’s market capitalization close to $4 trillion amid Wall Street’s growing confidence that Google can leverage its existing dominance in search, cloud infrastructure, and smartphone operating systems to roll out powerful new AI capabilities.
Koray Kavukcuoglu, Chief Technology Officer at DeepMind, informed the Financial Times that the large tech group has “significantly boosted performance” by training its AI models on Google’s custom-designed chips. “The ability to interface with consumers, customers, and businesses at this scale is something we can genuinely achieve through our deeply integrated approach,” he stated. He added that utilizing their proprietary silicon allowed Google to train Gemini 3 without depending on the costly Nvidia solutions relied upon by most other AI firms.
Some analysts suggest that OpenAI might have overreached in its pursuit of expansion at any cost. The startup has pledged to spend $1.4 trillion over the next eight years on compute power, securing massive procurement agreements with Nvidia, Oracle, AMD, and Broadcom. This projected expenditure dwarfs OpenAI’s current revenue, meaning its partners will likely need to secure loans to finance these joint ventures.
“This is a very, very high-stakes gamble for any company,” commented Sarah Myers West, Co-Executive Director of the nonprofit AI Now Institute.
The critical challenge currently facing OpenAI is identifying a revenue stream large enough to sustain such massive investments. Generating substantial advertising income via Sora, as Altman has suggested, will be difficult given the fierce competition already present from giants like Meta* and Alphabet. Furthermore, the startup is only just beginning to integrate advertising and shopping features into its core chatbot.
While Anthropic’s Claude chatbot isn’t as widely adopted by consumers as ChatGPT, The Financial Times notes that its sustained focus on AI safety has enabled the company to develop a more robust tool for enterprise clients, and its programming capabilities are considered best-in-class.
With over 800 million weekly users, OpenAI still leads in overall chatbot usage volume. However, according to analytics firm Similarweb, users are now spending more time interacting with Gemini than with ChatGPT. Simultaneously, Sensor Tower data confirms that ChatGPT maintains the top position among dedicated AI applications.
About the Author
