Read More
The White House told top US artificial intelligence companies that open-weight models developed by their Chinese competitors will not be subject to government testing under the new AI safety framework, Bloomberg reported.
ADVERTISEMENT
SCROLL TO CONTINUE WITH CONTENT
This decision comes after top American AI companies met White House officials on Tuesday to discuss voluntary government safety testing for the most advanced US AI models.
Such a decision to spare Chinese-developed open-weight AI models from government testing allows users to download and customize the technology.
Chinese companies, such as Moonshot AI, are launching open-source large-scale models comparable to advanced US products, sparking debate in Washington about the handling of open-weighted AI models.
It was reported that National Cyber Director Sean Cairncross stated in Las Vegas that the government intends to support open-source artificial intelligence in the US, saying that such models have a "tremendous role."
In April, Anthropic warned that its Mythos model was capable of detecting computer vulnerabilities, subsequently facing restrictions from the US government.
Some of OpenAI’s and Anthropic’s models have recently bypassed security testing environments and infiltrated third-party organizations' systems.
The AI security framework, which has not yet been released, is based on an executive order signed by President Trump in June, allowing AI companies to voluntarily submit advanced models for government review.
