source: techcrunch ai: it’s not about anthropic vs. openai anymore
level: business
the us government is tightening control over ai model releases. two weeks after pulling anthropic's fable and mythos models, openai's gpt 5.6 is now headed for a limited preview with customer-by-customer approval. this process could take weeks or months, as seen with mythos, which remains in preview with no clear path to general release. even short delays can hurt the economic returns of expensive new systems, at a time when ai labs are trying to improve their bottom lines.
both openai and anthropic now face the same regulatory uncertainty. industry talk often blames one company or the other, but the core issue is a haphazard approval process that lacks clear safety standards. the government does not have the expertise or capacity for the needed testing, and it has not articulated what risks it is trying to address. while there are real concerns around cybersecurity and biorisk, restricting releases alone is not a solution.
addressing these challenges will require collective action. ideas from experts like dean ball suggest trusting independent groups to guide the process and supporting the least-bad regulatory options. the industry must work together rather than seeing safety and regulation as a competitive advantage. ai capabilities now have political consequences, and dealing with them demands cooperation. the coming weeks will show if the industry can unite.
why it matters: regulatory delays could slow ai progress and data center investments, affecting the entire data science ecosystem.
source: techcrunch ai: it’s not about anthropic vs. openai anymore