Rapid advances in artificial intelligence are intensifying concerns about whether governments and technology companies can establish safeguards before increasingly powerful systems create serious economic, security or societal risks.
Warnings are now coming from executives at several leading AI developers, including Anthropic, OpenAI, xAI and Google. Some have called for slowing the development of the most advanced systems while governments develop testing and oversight standards.
Anthropic CEO Dario Amodei recently urged the industry to reduce the pace of development, warning that AI capabilities are accelerating rapidly. OpenAI CEO Sam Altman and xAI leader Elon Musk have also supported calls for greater caution, according to the Financial Times.
The warnings include the potential for advanced systems to facilitate cyberattacks, biological threats and other catastrophic harm. Other concerns involve job losses, misinformation, privacy, market concentration and the possibility that increasingly autonomous systems could act in unpredictable ways.
Washington remains deeply divided over how to respond.
President Donald Trump has emphasized rapid development, limited regulation and competition with China. His administration has generally favored voluntary testing and cooperation between companies and the federal government over mandatory restrictions. Trump argues that slowing development could allow another country to take the lead.
Some administration allies contend that companies concerned about safety can voluntarily slow their own work without requiring competitors to do the same. They also question whether leading firms are seeking regulations that could strengthen their market positions by making it harder for smaller companies to compete.
Congressional Democrats broadly agree that federal action is needed but remain divided over its scope. Sen. Bernie Sanders, an independent who caucuses with Democrats, and Rep. Greg Casar, D-Texas, support pausing advanced AI development and prohibiting the creation of artificial superintelligence.
Other lawmakers favor more limited oversight. Sen. Ruben Gallego, D-Ariz., has proposed a bipartisan congressional committee with subpoena authority, while Sen. Chris Van Hollen, D-Md., wants federal safety reviews of OpenAI’s models.
House Speaker Mike Johnson, R-La., has rejected an immediate moratorium but proposed a meeting among Trump, congressional leaders and AI executives to seek agreement on safety measures. He has said Congress should avoid adopting regulations before lawmakers understand the technology and potential consequences.
A narrower bipartisan approach could require developers to test advanced models for catastrophic risks and report the results. Sens. John Thune, R-S.D., and Amy Klobuchar, D-Minn., are pursuing such a proposal.
The Trump administration has already developed a voluntary process for reviewing some advanced systems before release, but the United States still lacks comprehensive federal AI legislation. Axios has reported that federal actions, state laws, and lawsuits are creating limited safeguards in the absence of a national framework.
Source: Washington Post, Axios, Financial Times