Beware putting the future of AI into the hands of politicians

Some overestimate the government's ability to develop novel solutions to challenging problems. NASA's spacefaring efforts were eventually surpassed by SpaceX, which has maintained a 100% safety record on human flight. The pandemic revealed a litany of federal failures before, during, and after, and the crisis. Yet, somehow, this same government supposedly holds the answers to keep us safe from AI. This is the belief currently driving many lawmakers. Take a new bill in Ohio that tasks the State Attorney General's office with licensing private Independent Verification Organizations to "independently verify" the safety of AI models. The AG would also appoint an advisory council to oversee this process. Consider the implications: Why would we want a politically motivated official or a digital luddite overseeing the direction of AI development? Attorneys general are wholly unqualified to "verify the safety" of this technology. Just as the private sector invented the three-point seatbelt, the private sector will develop the figurative "seatbelt" for AI — not bureaucrats with no technical expertise. Furthermore, these verification organizations are essentially unaccountable. Once chosen, they operate with no easy recourse for replacement. It is concerning that any government would encourage a private regulatory apparatus that lacks oversight and for which there is no market demand. Proponents claim the Ohio bill offers "voluntary" licensing, but this is disingenuous. The legislation provides a commercial liability shield to the verification organization's verified models, incentivizing developers to adopt recommendations regardless of whether they actually improve safety. States should not grant legal shields based on seals of approval from private entities that are ripe for regulatory capture. The claim that licensing is voluntary matters little given the government's proclivity for expanding its power. Whether recommendations are mandatory or not, the wisdom found in a politically connected commission is inherently suspect. Such a body could easily tilt the scales to damage the digital economy. For instance, what if a council is influenced by one specific player to kneecap its competitors? What if it prefers models aligned with a specific political ideology? Many remain outraged by biased content moderation following the 2016 election and during COVID; this legislation invites similar dangers by allowing AI policy to be dictated by unelected individuals. Beyond free speech concerns, allowing verification organizations to define "acceptable risk" creates a landscape of fluctuating rules. To the extent that AI is regulated, it should occur at the federal level to ensure consistency. This aligns with President Trump's National Policy Framework on AI, which notes that states should not regulate AI development because it is an inherently interstate phenomenon with national security implications. Lastly, there are serious intellectual property concerns at stake. If verification organizations are granted access to proprietary algorithms and models of AI players, and are auditing multiple competitors simultaneously, what is to stop them from revealing sensitive knowledge of either? The bill contains no provisions protecting trade secrets nor any recourse should there be a conflict of interest. If verification organizations were to gain access to the inner workings of leading AI models, will they have the cybersecurity protections and know-how to ensure they cannot be hacked by foreign adversaries like China? For these reasons, empowering State Attorneys General to chart the course for AI safety will only make the technology more prone to failure. One hopes these officials are humble enough to admit they lack the necessary technical skills. And if their prospective independently verified organizations possessed such expertise, they would be working in the industry rather than regulating it. Regardless of your stance on AI, putting its future in the hands of politicians is no way to fuel progress or safety. – Copyright 2026 Jon Decker, distributed by Cagle Cartoons newspaper syndicate. Jon Decker is the executive director of American Commitment and a senior fellow at the Parkview Institute.