According to CoinMeta, as reported by Benzinga, Nvidia CEO Jensen Huang argues that the risk of out-of-control AI should be regarded as an engineering issue, suggesting that powerful agents can be controlled through technical safeguards. In an interview with CNBC, he pointed out that the control of AI can be solved through engineering, but if assumptions are incorrect, the consequences could be unimaginable. He advocates against slowing down the development of AI and believes that companies that cannot safely control the system should not release it. Jensen Huang's comments come on the occasion of the launch of Nvidia, an open proxy security platform aimed at restricting access rights for autonomous AI proxies. He stated that the first step should be to deprive proxies of almost all permissions, granting access only when necessary. Currently, over 100 organizations are collaborating with this platform, including Microsoft, Anthropic, IBM, and others. Jensen Huang welcomes the use of OpenAI or the development based on Nvidia technology. Debates regarding the development of AI still exist; some AI leaders support slowing down cutting-edge development, while Jensen Huang believes that better engineering and monitoring are more practical solutions.