In Beijing with Donald Trump, the US Treasury Secretary said that the United States and China have discussed setting up safeguards for artificial intelligence (AI) to prevent it from being used for cyberattacks. In the race for AI dominance, both countries have put aside their rivalry to address the dangers of this technology. On Thursday, US Treasury Secretary Scott Bessent stated that the two countries have exchanged ideas on establishing safeguards for AI. “We will discuss AI safeguards with the Chinese,” the Secretary told American media CNBC from China. He noted that Beijing has a highly advanced AI industry but lags significantly behind the United States.
Secretary Bessent, part of the delegation accompanying President Trump to Beijing, emphasized the importance of not stifling innovation. He mentioned that both countries, sharing concerns over the potential cybersecurity threats posed by AI, will develop a ‘protocol’ defining the path forward in AI matters. This protocol will help ensure that non-state actors do not misuse these AI models. Despite this, he reassured CNBC that the US government does not want to hinder innovation and expressed satisfaction with how major tech companies have implemented these models so far.
Last week, the International Monetary Fund (IMF) called for enhanced international cooperation regarding AI due to the risks of cyberattacks using AI tools. This issue was particularly highlighted by the introduction of the startup Anthropic’s new model, Mythos, which identified “thousands” of ‘zero-day’ vulnerabilities in online-accessible programs during internal testing. A ‘zero-day’ vulnerability refers to a weakness that designers and users are unaware of. Without intervention to address these vulnerabilities, hackers could potentially exploit them for cyberattacks.
