The competition for artificial intelligence (AI) has entered a new\nphase. Amid allegations from the United States government that Chinese\ncompanies are stealing US-made AI model capabilities, Nvidia CEO Jensen\nHuang has taken a different position.<\/p>\n
Huang rejected the notion that the practice of AI model distillation\nis identical to theft. According to him, studying competitors\u2019\ntechnology is a common part of competition that occurs across various\nindustries.<\/p>\n
\u201cThat is what competition is,\u201d said Huang when questioned during the\nSquable Box programme, as reported by CNBC on Monday, 28 September\n2026.<\/p>\n
Huang\u2019s statement has reignited the debate regarding the boundary\nbetween technological learning and the theft of intellectual property\nwithin the AI industry. This issue is becoming increasingly significant\nbecause American AI companies spend vast sums of money training models,\nwhile competitors can utilise the outputs of those models to train\ncheaper systems.<\/p>\n
This practice is known as distillation. Technically, distillation is\na legitimate method for transferring the capabilities of large-scale AI\nmodels into smaller models.<\/p>\n
The problem arises when this method is used against a competitor\u2019s\nmodels without permission. The US government believes such practices\ncould potentially serve as a shortcut to acquiring capabilities built\nwith billions of US dollars in investment.<\/p>\n
Huang views it from a different perspective. He stated that companies\nessentially have the right to test and study competitors\u2019 products.<\/p>\n
Nvidia itself, Huang noted, experiences similar situations. The\ncompany\u2019s hardware products are frequently purchased by competitors,\ndismantled, and studied to understand the technology used within\nthem.<\/p>\n
He admitted that it would certainly be more comfortable if\ncompetitors did not do this. However, according to Huang, competition\nactually encourages companies to continuously improve their\ntechnology.<\/p>\n
If a company does not want its products to be used in a certain way,\nHuang continued, the company can tighten customer selection or terminate\nservice access.<\/p>\n
This view stands in opposition to US Treasury Secretary Scott\nBessent. In July, Bessent stated that the government had found traces of\nAmerican models within several Chinese AI models.<\/p>\n
Bessent even opened the possibility of imposing sanctions. \u201cWe have\nthe ability to impose sanctions on them because of this theft,\u201d he told\nFox Business, as quoted by CNBC.<\/p>\n
Huang has not changed his stance. A day after Bessent\u2019s statement, he\nsaid that US companies should still be permitted to use AI models made\nin China.<\/p>\n
Accusations Intensify<\/strong><\/p>\n The debate shifted from a mere war of words to a national security\nissue in September. Three US government agencies accused six Chinese AI\ndevelopers of conducting what is described as an \u201cindustrial-scale\nknowledge distillation campaign.\u201d<\/p>\n The six developers include Alibaba and DeepSeek. The allegations were\nbrought forward by the National Security Agency (NSA), the Federal\nBureau of Investigation (FBI), and the Cybersecurity and Infrastructure\nSecurity Agency (CISA).<\/p>",
"url": "https:\/\/jawawa.id\/newsitem\/as-us-threatens-sanctions-on-china-nvidia-ceo-defends-ai-distillation-1790733393",
"image": ""
},
"sponsor": "Okusi Associates",
"sponsor_url": "https:\/\/okusiassociates.com"
}