San Francisco, California — The emergence of agentic artificial intelligence has become a focal point in discussions about the future of technology this year. Advocates believe that these self-operating AI systems could dramatically transform various sectors, including retail, healthcare, and many aspects of daily life. However, concerns linger about the current capabilities of such technologies and our readiness to integrate them widely.
While the term “agentic AI” might be relatively new, the associated risks and challenges are well recognized. It is impractical to create entirely new legal frameworks for every advancement in AI technology if lawmakers aim for coherence and functionality across diverse jurisdictions. Instead, focus needs to shift toward applying existing laws to manage the potential threats posed by agentic AI.
Agentic AI encompasses systems that undertake tasks and pursue objectives either on their own or with minimal human oversight. Although it signifies a new phase in technology, agentic AI builds on decades of innovations aimed at automating processes previously managed without digital assistance. With large-scale autonomous computing already in practice, the potential dangers increase if these systems lack thoughtful design and careful implementation.
The rise of agentic AI raises new risks due to the reduced role of human judgment, which can lead to unpredictable and potentially harmful outcomes. Lawmakers and regulatory bodies must concentrate on current legal frameworks that can address these emerging concerns.
Many existing laws, including those protecting consumer rights and privacy, continue to be applicable to the risks associated with agentic AI. These regulations cover a variety of issues, such as unfair business practices, data protection, and intellectual property rights. Several states, including California, Colorado, Utah, and Texas, have taken steps to specifically regulate AI technology. Additionally, numerous states have adopted comprehensive privacy laws that give consumers established rights regarding the use of their personal data, particularly in critical areas such as finance, healthcare, and employment.
Though these laws do not explicitly mention “agentic AI,” they include provisions that tackle related risks, including algorithmic biases, a lack of transparency in automated decision-making, and the potential for consumers to be misled into thinking they are interacting with a human instead of a machine.
A significant concern persists about accountability when agentic AI causes harm, especially since its development often involves multiple parties. Responsibility tends to be determined through contractual agreements, but existing laws outline harm and risk while not clearly addressing who bears that burden.
As various stakeholders—from developers to users—are part of the agentic AI framework, the distribution of risk continues to be a contentious topic. How these stakeholders negotiate and manage responsibility remains crucial, particularly given the complex and opaque nature of many AI technologies. For instance, if an AI tool incorrectly purchases items outside of a consumer’s instructions, determining liability—be it the consumer, the merchant, or the AI developer—can be complicated.
Such scenarios emphasize the need for robust contracts among all parties involved to delineate responsibilities and rights. Regulatory agencies also have the authority to promote protective terms in these agreements or establish rigorous internal processes for evaluating contractual relationships.
Given the rapid pace of technological evolution, creating new laws specifically for agentic AI may not be the most effective approach. Instead, the pressing issues surrounding the allocation of existing legal risks will likely surface in private negotiations rather than through legislative action.
This article was automatically written by Open AI. The people, facts, circumstances, and story may be inaccurate, and any article can be requested removed, retracted, or corrected by writing an email to contact@publiclawlibrary.org.