AI economy driven by frontier labs, open models, and applications

The artificial intelligence economy is characterized by three competing forces: closed-source frontier labs, open-weight models primarily from China, and application companies. While frontier labs like OpenAI and Anthropic have experienced significant revenue growth and high demand, they face increasing pressure regarding return on investment (ROI).

Estimates suggest AI spending currently accounts for between 0.5 and 1 percent of all white-collar salaries in the United States. This scale has led to scrutiny, including concerns from Palantir’s Alex Karp regarding enterprises spending heavily on tokens without corresponding productivity gains.

Competition is intensifying as Meta and xAI release increasingly capable models to compete with Google, OpenAI, and Anthropic. Simultaneously, Chinese companies such as Zhipu and Moonshot are gaining momentum with open models like GLM 5.2 and Kimi K3, which perform near frontier levels at a fraction of the cost.