Journalist Karen Hao described OpenAI as an empire driven by quasi-religious fervor that threatens democratic governance and fosters AI-driven colonialism [1].
Hao's warnings come as the company continues to expand its global influence, suggesting that the pursuit of artificial intelligence dominance is no longer just a business venture but a systemic risk to sovereignty.
Speaking during a July 3, 2026, broadcast for a July Fourth special on Democracy Now!, Hao said that the development of AI in Silicon Valley has been shaped by "quasi-religious movements" [1]. She said that this zeal fuels the growth of OpenAI under CEO Sam Altman, framing the organization as a modern-day empire [2].
Hao has reported on OpenAI since 2018 [2]. She said that the company's trajectory shifted rapidly after ChatGPT launched in 2022, when it became the fastest-growing app in history [4]. This rapid scaling has concentrated immense political and economic power within a single private entity.
According to Hao, this concentration of power creates a new form of colonialism, where the infrastructure and logic of AI are imposed on a global scale [1]. While she has previously noted that everyone has a stake in the development of AI [3], she now emphasizes the danger of that development being steered by a small, ideologically driven group.
These concerns arrive at a critical juncture for the company. OpenAI is expected to release the latest version of ChatGPT later this month [4]. The upcoming release is seen as another step in the company's effort to maintain its lead in the AI race, a race Hao suggests is being run with a fervor that bypasses traditional democratic checks.
“OpenAI is an empire”
The characterization of OpenAI as an 'empire' shifts the conversation from technical safety to political power. By framing the company's growth as a form of colonialism, Hao suggests that the risks of AI are not just algorithmic errors, but the erosion of national autonomy and democratic oversight in favor of a centralized, corporate-led intelligence infrastructure.



