US States Pursue AI Regulation Amid Global Treaty Debate

The United States is currently navigating a complex landscape of artificial intelligence (AI) regulation, with individual states like California and New York taking proactive steps to establish governance frameworks. This state-led initiative contrasts with the ongoing international discussions and aspirations for a unified global treaty to manage AI risks. The absence of a comprehensive federal AI law in the U.S. has created a vacuum, prompting sub-national entities to develop their own approaches, potentially leading to a patchwork of regulations across the country.

California, a hub for technological innovation, has been particularly active in exploring AI governance. In September 2023, Governor Gavin Newsom signed an executive order aimed at establishing a framework for AI development and deployment within the state. This order directed state agencies to develop guidelines and best practices for AI use, focusing on safety, equity, and accountability. The state's legislative efforts have also included proposals to address AI's impact on employment, bias in algorithms, and the responsible use of AI in critical sectors such as law enforcement and healthcare. These initiatives reflect a growing recognition of the need to balance innovation with risk mitigation.

New York has also emerged as a significant player in AI regulation. In January 2024, the New York City Council passed legislation requiring businesses to conduct bias audits for AI hiring tools before their deployment. This law, which took effect in July 2024, mandates that employers using AI in their hiring processes must have these tools audited for discriminatory effects. The legislation aims to protect job applicants from unfair practices and promote equitable hiring. Furthermore, New York State has been considering broader AI legislation, including proposals that would require transparency in AI systems and establish an AI task force to advise on regulatory strategies. The state's focus on specific applications of AI, such as in employment, highlights a pragmatic approach to addressing immediate concerns.

These state-level actions underscore the challenges in achieving a cohesive national or international AI regulatory strategy. While a global treaty is often cited as the ideal solution for managing AI's far-reaching implications, the practicalities of international consensus-building are immense. The differing priorities, technological capacities, and ethical considerations among nations make a singular treaty a distant prospect. Consequently, the current environment favors a more decentralized and adaptive regulatory model. The divergence in approaches between California and New York, for instance, suggests that even within a single country, a unified regulatory vision may be difficult to attain. This fragmentation could create compliance challenges for companies operating across different jurisdictions and may necessitate ongoing efforts to harmonize disparate rules.