Twenty-five technology companies, including Nvidia, Meta and Microsoft, have published an appeal "Open Weights and American AI Leadership", urging regulators to refrain from imposing restrictions on artificial intelligence models with open weights. The authors of the document warn that such bans will not strengthen leadership in the field of AI, but rather lead to the monopolization of the market by several closed developers. The signatories draw a parallel with the free software movement of the 1980s, emphasizing that technological excellence depends on the ability to build an open ecosystem for various industries, rather than on control over a single advanced proprietary model. In addition, the authors of the appeal defended the practice of distilling models, suggesting that controversial issues should be resolved through targeted legal and commercial mechanisms, rather than through total bans.

The initiative has received public support from Nvidia CEO Jensen Huang and Microsoft CEO Satya Nadella. Huang noted that open models enhance cybersecurity, accelerate the spread of innovation, and strengthen technological sovereignty. It is noteworthy that major developers such as OpenAI and Anthropic are not among the signatories. This appeal appeared against the background of an escalating discussion on the regulation of open technologies, where possible restrictions on the use of such foreign-made products are being discussed.

The practical value of open algorithms was recently demonstrated by an incident on the Hugging Face platform. The company's specialists, who were trying to use advanced commercial neural networks to analyze 17,000 cyberattack log entries, encountered request blocking: built-in security filters could not distinguish a legitimate investigation from an attempt to get hacking help by responding to malicious commands and exploit texts.