Sam Altman says there was a big reason OpenAI released its open-weight models

In a recent media briefing, Sam Altman, CEO of OpenAI, revealed that the rise of China's open-source AI models significantly influenced the decision to release its own open-weight models this month. "It was clear that if we didn't act, the landscape would be dominated by Chinese open-source solutions," Altman stated, emphasizing that this consideration played a crucial role in their strategy, alongside other factors. On August 5, OpenAI unveiled two new open-weight models: gpt-oss-120b and gpt-oss-20b. The former is tailored for advanced computing systems, while the latter can operate on most standard desktops and laptops. This launch signifies OpenAI's first foray into open-weight models since the introduction of GPT-2 in 2019, marking a shift from their closed model approach. Unlike the proprietary models such as GPT-3, GPT-4, and the newly announced GPT-5, which are restricted in usage, open-weight models allow users to run and customize them locally. Altman proclaimed on social media that OpenAI's gpt-oss represents the "best and most usable open model in the world," aligning with the organization's mission to foster artificial general intelligence (AGI) that benefits humanity. Altman expressed optimism about the potential for an open AI ecosystem in the U.S., built on democratic principles and accessible to all. This commitment contrasts with other tech giants; for instance, Meta has also embraced open-weight models but remains cautious about its open-source strategies. The competitive landscape has intensified with the emergence of companies like DeepSeek, whose R1 model has gained attention for its performance and affordability, prompting discussions about the future direction of AI technology in America. Altman acknowledged this shift in a January AMA, stating that OpenAI needed to reevaluate its approach to open-source AI in light of these developments.