Anthropic, Meta, Google and OpenAI released AI models one after another within a week, adding to what is being called “model fatigue” in corporate settings, CNBC reported on Saturday.

Frequent updates are forcing CEOs and IT managers to spend more time and resources comparing costs and performance, it said.

This week began with Anthropic unveiling Claude Fable 5.1 and Claude Mythos 5.1. Meta announced Muse Spark 1.3, and Google unveiled Gemini 3.8 Flash. OpenAI released GPT-6 Astra, highlighting cybersecurity and computer-use capabilities.

Nvidia reached an official agreement to acquire open-source AI platform Hugging Face for $12.9 billion. Last month, it also unveiled Nemotron 3.5 Lightning, which can run on a single GPU in a laptop or desktop computer.

The industry points to competition for market share as a driver of the speed race. Ahmed Abbasi (아메드 아바시) of the University of Notre Dame’s Mendoza College of Business said major developers are competing to avoid falling behind each other and want to show developers the pace of innovation. He described it as a battle for “share of wallet.”

Gartner forecast AI spending this year at $2.59 trillion, up 47 percent from 2025. More than half will go to infrastructure, but more than $1 trillion is expected to be spent on services, software, cybersecurity, models and other tools.

The burden is also rising in the field. Jin Ryu (젠 루), chief executive of AI startup Runpod, said, “Model fatigue really exists,” adding that innovation is continuing but there is also a lot of noise aimed at standing out. Noah Faro (노아 파로), chief technology officer of AI finance startup Parsight, said it may not be a coincidence that major companies announced updates in the same week.

Companies say it is difficult to verify all the updates pouring out. Suresh Basudevan (수레시 바수데반), chief executive of Clockwork Systems, said that even if he has to evaluate 10 AI models for a specific task, in practice he can only narrow it down to 5. He said it is very difficult to evaluate every model that comes out, one by one.