The 118th US Congress commenced its session on January 3, 2023, marking a year since its inception. During that period, I traveled to Las Vegas to attend the Consumer Electronics Show (CES) alongside several other Congressional technology policy staff members. While engrossed in the latest sensors and AI-enabled devices, we pondered the implications of the House Speaker vote broadcasted on C-SPAN for democracy and, more specifically, for technology policy.
Reflecting on the extensive advocacy efforts aimed at passing bills safeguarding data and enhancing competition, it was evident that minimal progress had been achieved on these fronts within the 118th Congress. Despite the technological marvels surrounding us, the real surprise came a few weeks later when the dialogue between columnist Kevin Roose and Bing’s Sydney chatbot made headlines in the New York Times in February 2023. This event marked a significant shift in how technology journalism resonated with lawmakers, capturing their attention after years of congressional service.
Subsequent to Roose’s account, numerous media outlets highlighted Congress’s perceived lag in AI advancement. However, the reality was more nuanced. With the intricate landscape of AI risks, Congress faced a myriad of proposals, including landmark bills from the 117th Congress, addressing generative AI concerns. In April, I commenced monitoring national legislative initiatives related to conceptual AI in collaboration with Tech Policy Press and colleagues.
Fast forward to the present, the landscape of proposed legislation has evolved significantly. The focus of 2023 centered on elucidating how bills like the Platform Accountability and Transparency Act (PATA) (S. 1876), American Data Privacy Protection Act (H.R. 8152 117th), and the American Choice and Innovation Online Act (2033) could equip regulators to tackle diverse generative AI risks. Amidst this discourse, the Senate Majority Leader advocated for an innovative customer relationship approach, emphasizing the need to address the challenges posed by conceptual AI through unconventional means.
While incremental progress was observed, the reemergence of pivotal competition legislation from the previous Congress and the endorsement of current opposition reform bills underscored a semblance of advancement. Notably, the House Energy and Commerce Committee faced hurdles in advancing the ADPPA due to its intricate nature, necessitating new members to acclimate to its complexities. Concurrently, the discourse surrounding the integration of public data into training data using tools like ChatGPT raised pertinent questions about data handling ethics.
The reintroduction of PATA in June 2023 and the prominence of the Kids Online Safety Act (S. 1409) in discussions underscored the ongoing deliberations on social media regulation. Lawmakers acknowledged the pivotal role of social media platforms in disseminating generative AI-generated content, signaling the need for stringent oversight.
Looking ahead to 2024, the focus remains on fostering viable policies that underscore the significance of opposition reform, data security, and content moderation transparency. As the fervor surrounding stakeholder engagements subsides, the emphasis shifts towards substantive legislative actions that align with the core tenets of digital market oversight.
In parallel, Congress is striving to establish a framework that enables individuals to differentiate between authentic content and AI-generated information. Proposals such as the DEEP FAKES Accountability Act (H.R. 3230) and the AI Labeling Act (S. 2691) mandate disclosures for AI-generated content, emphasizing the need for standardized practices to enhance user discernment. The enactment of the “Generative AI Detection and Watermark Competition” under Section 1543 of the 2024 National Defense Authorization Act underscores the commitment to promoting research on technical standards essential for disclosures.
Moreover, Congress is engaged in deliberations on the “hub-and-spoke” approach to evaluating the benefits and risks of AI deployment. Various proposals advocate for comprehensive risk assessments and countermeasures tailored to both the technology itself and its application across diverse sectors. Initiatives like the Algorithm Accountability Act (AAA) (S. 2892) and the Artificial Intelligence Research, Innovation, and Accountability Act (AIRIA) (S. 3312) underscore the imperative of establishing robust oversight mechanisms to mitigate AI-related risks effectively.
Furthermore, the integration of security requirements through procurement processes signifies a strategic approach to ensure the responsible development and utilization of AI systems. Legislative efforts such as the AI LEAD Act (S. 2293) and the Federal Artificial Intelligence Risk Management Act (S. 3205) underscore the importance of rigorous evaluation and documentation of AI systems procured by government agencies.
As we navigate through 2024, the emphasis remains on securing government funding and establishing standardized frameworks to guide AI assessments, disclosures, and monitoring practices. The collaboration between NIST and relevant stakeholders is crucial in setting the foundation for consistent and effective AI governance practices.
In conclusion, the legislative landscape in 2024 reflects a concerted effort to address the complexities of AI regulation and oversight. By prioritizing research, standardization, and ethical considerations, Congress aims to foster a regulatory environment that balances innovation with accountability. As we anticipate further developments in the realm of AI policy, the commitment to advancing these critical issues remains steadfast.
