In the evolving landscape of artificial intelligence regulation, the insights of Geoffrey Hinton, a pioneer in the field and a 2024 Nobel Prize laureate in Physics, shine a light on the path forward. During his impactful presentation at the National Conference of State Legislatures (NCSL) Annual Summit held in Chicago, Hinton argued for a paradigm shift in how AI developers approach the creation of large language models. He proposed that these technologies should be “intrinsically good,” akin to the moral education we strive to instill in children.
Hinton’s assertion that AI products, such as chatbots, ought to be trained to exhibit good behavior underscores a critical perspective: that the responsibility for ethical AI lies not just with regulators but with developers themselves. As he articulated during the conference, “You can model good behavior,” suggesting that intentional training could yield AI that aligns more closely with societal values. This approach challenges current regulatory practices, which Hinton criticized as “guardrail” measures that often enrich lawyers while exacerbating conflicts between state and federal jurisdictions.
A pivotal moment in California’s legislative efforts to establish such standards came with House Bill 1047, which aimed to set a framework for ethical AI but was ultimately vetoed by Governor Gavin Newsom. Hinton expressed confusion over the veto, suggesting that deeper motivations might be at play. The disconnect between legislative intent and executive action raises questions about the efficacy of existing governance structures in addressing the rapid advancements in AI technology.
Moreover, Hinton emphasized the need for transparency in AI development, advocating that developers should disclose testing processes and results to both local and federal governments before launching large language models. This transparency would not only enhance accountability but also help mitigate risks associated with AI technologies that may inadvertently perpetuate harmful behaviors.
The urgency of Hinton’s message resonates within a broader context. With AI technologies increasingly integrated into business, government, and society, the pace of legislative response has been remarkable. Over the past three years, nearly 100 AI-related bills have been adopted across all states and territories, addressing issues from copyright ownership of AI-generated content to safeguarding critical infrastructure from AI vulnerabilities. According to Utah Representative Paul Cutler, who moderated a session at the summit, the legislative landscape is evolving faster than many historical climate shifts, highlighting the need for a balanced approach that considers both innovation and societal risks.
However, the regulatory environment faces significant challenges, particularly from federal entities. The Trump administration’s December 2025 executive order established an AI Litigation Task Force, aiming to challenge state regulations deemed overly burdensome. This federal posture reflects a broader narrative that prioritizes rapid innovation, often at the expense of comprehensive oversight. Hinton cautioned against this perspective, arguing that an absence of regulation could lead to detrimental outcomes, especially if AI systems are trained on unfiltered data that may normalize harmful behaviors.
Hinton’s analogy of training children to recognize right from wrong serves as a compelling metaphor for responsible AI development. He stated, “Your controls over it are much the same as your controls over your children,” emphasizing the importance of reinforcing positive behaviors. This perspective aligns with recent studies suggesting that ethical training in AI can lead to better alignment with human values, ultimately fostering technologies that enhance societal well-being rather than undermine it.
As we look toward the future, the conversation surrounding AI regulation must evolve to reflect the complexities of the technology itself. The interplay between innovation and regulation is not a binary choice; rather, it is a dynamic relationship that can drive progress in beneficial directions. By focusing on the intrinsic goodness of AI and fostering a collaborative environment between developers and regulators, we can navigate the challenges ahead while ensuring that technology serves humanity’s best interests. The stakes are high, and the choices made today will shape the ethical landscape of AI for generations to come.
Reviewed by: News Desk
Edited with AI assistance + Human research
