California SB 1047
About the author: Tomas Corza is a marketing strategist, Jobs-to-be-Done practitioner and musician.
Click here to contact Tomas Corza.
Artificial Intelligence (AI) is no longer a futuristic concept confined to science fiction, it’s here, shaping industries, driving innovation, and transforming society. As this technology evolves at an unprecedented pace, governments and policymakers are feeling an increasing urgency to establish regulations. The question is: who is leading these efforts, what are they trying to achieve, and why is AI regulation such a pressing issue? Let’s break this down.
Disclaimer:
The information contained in this site is provided for informational purposes only, and should not be construed as legal advice.
What is SB 1047?
SB 1047, introduced by California State Senator Scott Wiener, is a bill designed to regulate AI at the model level. This is a controversial move because it attempts to govern AI as a general-purpose technology, instead of focusing on specific applications like hiring algorithms, self-driving cars, or medical devices. The bill seeks to impose sweeping regulatory controls on the development and deployment of AI models, which are the underlying systems that drive AI-powered innovations.
The core idea behind the bill is to ensure that AI technologies are safe, ethical, and transparent before they can be widely used. By regulating the AI models themselves, SB 1047 aims to provide oversight from the ground up, preventing potential harms before they manifest AI powered products and services.
Key Provisions of SB 1047
Here are some of the most important aspects of SB 1047:
Third-Party Audits: SB 1047 requires AI companies to submit their models for third-party audits. These audits would assess the risks, biases, and safety of the models before they can be deployed on a large scale. This means that a third-party organization, separate from the developers themselves, would review and evaluate AI models to ensure they meet certain safety and ethical standards.
Model-Level Regulation: Unlike other AI regulations that focus on specific use cases (like facial recognition or autonomous vehicles), SB 1047 proposes regulating general-purpose AI models, the foundational systems that can be adapted to perform a wide range of tasks. This makes the bill unique and more far-reaching, as it tackles the core technology behind AI, not just its end-use applications.
Mandatory Reporting: AI developers would need to disclose information about their models, including how they were trained, what data was used, and how they will be implemented. This reporting would provide transparency into how AI systems are being created and deployed, which is meant to address concerns about AI models potentially being biased or harmful.
Precautionary Principle: SB 1047 follows the precautionary principle—the idea that new technologies should be regulated upfront to prevent harm, rather than allowing them to develop freely and regulating them only after problems arise. This contrasts with the U.S. tradition of "permissionless innovation," where new technologies can be developed and deployed with minimal oversight until real-world issues are detected.
Risk Assessment and Liability: AI companies would be required to perform detailed risk assessments for their models, identifying potential dangers such as bias, misuse, or malfunction. In cases where the models are found to cause harm, developers could be held liable for damages. This raises the stakes for AI development, especially for smaller companies and startups that may not have the resources to handle extensive legal and compliance requirements.
Alignment with European Approaches: SB 1047 takes inspiration from the European Union’s AI Act, which also seeks to impose stringent controls on AI technologies. California’s bill reflects a similar desire to lead in AI regulation, with Senator Wiener stating that if Congress does not act, California should step up as the de facto regulator of AI in the U.S.
Why is SB 1047 Controversial?
The introduction of SB 1047 has sparked intense debate among tech companies, policymakers, and experts in the field of AI. Here are some of the major points of contention:
Impact on Innovation: Critics argue that regulating AI at the model level could stifle innovation. Developing AI models is a complex, iterative process, and adding regulatory hurdles could slow down progress. Smaller companies, in particular, may struggle to comply with the bill’s demands, giving an advantage to larger tech giants that have the resources to handle compliance. There's a concern that California could become less competitive as a tech hub if its regulations are too stringent.
Overreach of State Law: Some critics worry that California is overstepping by attempting to impose regulations on technologies that have national and global implications. AI development often spans borders, and regulating it at the state level may create a patchwork of conflicting regulations across the U.S., making it difficult for companies to operate efficiently. Federal lawmakers, including figures like Nancy Pelosi and Zoe Lofgren, have expressed concern that AI regulation should happen at the federal level to avoid this issue.
Global Competitiveness: There’s also concern about how these regulations could impact the global AI race, particularly with competitors like China. Some fear that if California imposes strict rules, it could drive AI innovation out of the state—or even out of the U.S.—into countries with more relaxed regulations. With China aggressively investing in AI, the U.S. could lose its competitive edge if the regulations hinder progress.
Liability Issues: The liability provisions of SB 1047 are another source of controversy. AI developers could be held legally responsible for the actions of their models, even if those models are used in ways the developers did not intend. This has led to concerns about open-source AI models, which are freely available for others to use and modify. If developers are held responsible for how these open-source models are used, it could discourage the development and sharing of open-source AI, which has been a major driver of innovation in the field.
Support for SB 1047
Despite the criticisms, many believe that AI regulation is necessary and that California is in a position to lead the way. Supporters argue that the risks posed by AI—such as biased decision-making, privacy violations, and even potential threats to democracy—are too great to leave unchecked. They believe that transparent, ethical AI development will ultimately benefit society and prevent harm before it happens.
Moreover, supporters of the bill argue that federal inaction has left a regulatory void that states like California need to fill. While federal legislation has been slow to materialize, states have traditionally been laboratories for new ideas, and California has often been a trendsetter when it comes to tech policy.
What’s Next for SB 1047?
As of now (9/10/2024), SB 1047 is awaiting a decision from California Governor Gavin Newsom, who will either sign the bill into law or veto it. Many are watching closely to see what happens, as the decision could set a precedent for how AI is regulated across the U.S. and even internationally.
If passed, SB 1047 could have wide-reaching consequences for the tech industry, potentially reshaping how AI is developed and deployed not only in California but across the country. Whether you see it as a necessary step to protect society or a dangerous move that could slow down innovation, SB 1047 is a significant piece of legislation that will shape the future of AI regulation.
SB 1047 is a bold attempt by California to regulate AI at its core. It reflects growing concerns about the impact of AI on society and the desire to prevent harm before it occurs. But whether this precautionary approach will foster safer AI or stifle innovation remains a heated debate.
