With science in flux, AI safety is a moving target

The near constant evolution of artificial intelligence (AI) is creating unique challenges for policymakers seeking to establish safety protocols and regulatory guidelines in the field.

Current policy challenges: The U.S. Artificial Intelligence Safety Institute faces significant hurdles in recommending concrete safeguards for AI systems due to the technology’s rapidly evolving nature.

• Elizabeth Kelly, director of the U.S. AI Safety Institute, highlighted the difficulty in establishing best practices when the effectiveness of various safeguards remains uncertain
• Cybersecurity concerns are particularly pressing, with AI systems vulnerable to “jailbreaks” – methods that bypass established security measures
• The manipulation of digital watermarks meant to identify AI-generated content presents another significant challenge for regulatory guidance

Collaborative approach: The Biden administration’s AI Safety Institute is working to address these challenges through partnerships with various stakeholders.

• The institute maintains active collaborations with academic institutions, industry players, and civil society organizations
• These partnerships inform the institute’s technical evaluations and safety recommendations
• Recent initiatives included the first global gathering of AI safety institutes in San Francisco, bringing together representatives from 10 countries

International coordination: The inaugural meeting of global AI safety institutes marked a significant step toward developing standardized safety protocols.

• The gathering focused on creating interoperable safety tests across different countries
• The meeting notably featured more technical experts than typical diplomatic assemblies
• Kelly described the atmosphere as “getting the nerds in the room,” emphasizing the technical nature of the discussions

Political considerations: The future of AI safety regulation spans political divisions and administration changes.

• Kelly emphasized that AI safety is a “fundamentally bipartisan issue”
• This characterization suggests the institute’s work may continue regardless of potential administration changes
• The comment came in response to questions about the institute’s future under different political leadership

Looking ahead: The complexity of AI safety regulation highlights the need for flexible, evolving approaches that can adapt to rapid technological changes while maintaining robust security measures.