It’s not about Anthropic vs. OpenAI anymore

The U.S. government is asserting direct control over the release of frontier artificial intelligence models, placing OpenAI's latest system, GPT 5.6, into a regulatory holding pattern just two weeks after a similar intervention halted the wide release of models from rival Anthropic [1]. The Information reported Thursday that GPT 5.6 will be released only into a limited preview, with the government approving access on a "customer by customer" basis until a general release can be authorized [1]. OpenAI executive Sam Altman reportedly projected the preview could last a "couple of weeks," but the timeline remains uncertain [1]. Anthropic's Mythos model has already been confined to a preview period for months with no clear path to a full launch [1]. The new scrutiny arrives during a period of exponential investment in AI, driven by the rapid scaling and public release of large language models [2]. The current boom, initiated by the development of the transformer architecture in 2017, has seen models like ChatGPT integrated across sectors [2]. The dominant firms in the U.S. technology industry, often referred to as Big Tech, now include companies with market capitalizations making up roughly a quarter of the S&P 500 [3]. Dean Ball, a fellow at George Mason University and a soon-to-be OpenAI employee, wrote that it is "not clear what kind of safety assurances could be put in place to satisfy regulators" [1]. Ball argued that the U.S. government lacks the expertise and capacity for the necessary testing, and that no formal articulation of the specific risks concerning regulators has been made [1]. The government's move echoes historical cycles of funding and restriction in the field. The history of artificial intelligence shows that after initial enthusiasm and millions of dollars in U.S. government funding in the 1950s and 1960s, criticism and congressional pressure led to a halt in undirected research in 1974, a period later known as an "AI winter" [2]. Industry dynamics are further complicated by the financial stakes. The primary report notes that many prominent people in the industry have "billions of dollars" riding on one company or the other [1]. The trial and conviction of FTX founder Sam Bankman-Fried in 2023, a case that raised broad awareness of criminal activity in emerging tech markets, underscored the high financial and regulatory risks present in novel technology sectors [4]. The current impasse leaves both OpenAI and Anthropic facing the same regulatory barrier, a situation the primary report argues requires collective industry action rather than competitive maneuvering [1]. The report concludes that AI models have progressed to a point where their capabilities carry real political consequences, and addressing them will demand a unified approach [1].