China’s Ministry of State Security has issued a warning over the potential risks artificial intelligence poses to the country’s political and social security, highlighting growing concerns in Beijing about the technology’s impact. In its first public statement addressing AI, the agency outlined threats including the misuse of generative AI tools such as deepfake media, AI-generated text, and “intelligent troll armies” to fabricate political rumors, spread harmful information, and foment social divisions. Chen Yixin, China’s state security minister, described these actions as forms of “cognitive warfare” that endanger the nation’s political, institutional, and ideological security.

This heightened rhetoric marks a shift in China’s public discourse on AI safety, traditionally led by the Cyberspace Administration of China with an emphasis on regulatory and technical matters. While Beijing’s statement did not echo the more extreme warnings from some US researchers about AI potentially causing human extinction, it reflects growing unease about the technology’s strategic vulnerabilities, especially given advances by companies such as OpenAI and Anthropic. Chinese officials have reportedly raised these concerns in recent discussions with US counterparts, cautioning against moves to ban open-source Chinese AI models. Tensions over AI are expected to feature in the forthcoming meeting between US President Donald Trump and Chinese President Xi Jinping.

Meanwhile, in the United States, voices across the political spectrum have recently converged on the issue of AI safety. Veteran socialist Senator Bernie Sanders and right-wing activist Steve Bannon are set to jointly advocate for stricter regulatory measures on AI at a “Pro-Human Assembly” event in Washington. This unusual alliance reflects a broader bipartisan push to address potential societal harms linked to rapid AI development. Among invited speakers are figures from both major parties, including Republican Senator Marsha Blackburn and progressive Congressman Greg Casar, emphasizing the wide-ranging concern.

The event occurs amid calls from leaders of major AI companies such as OpenAI, xAI, and Anthropic, who have urged a coordinated slowdown in developing advanced AI models. Public opinion polls indicate significant apprehension among Americans; a recent NBC survey found that about two-thirds perceive AI as a high or very high risk to society. Proposals gaining traction include bans on new AI data centers and enhanced government oversight of AI models, which currently undergo only voluntary reviews.

Despite this growing consensus, President Trump has resisted regulatory efforts, describing them as part of a “SICK conspiracy” aimed at hindering AI and data center development, which he claims primarily benefits China. Trump has asserted that AI requires no special guardrails beyond effective presidential leadership. His vice president, JD Vance, has taken a more moderate stance, suggesting that some regulation is needed through bipartisan cooperation.

The expanding debate over AI safety reflects rising global concerns about the technology’s rapid evolution and its implications for security, governance, and society. Both China and the United States are grappling with how to balance innovation with protective measures amid uncertainty about AI’s long-term impacts.