TL;DR

State-level AI legislation in the US is filling a regulatory vacuum left by federal inaction. These laws, though sometimes fragmented, are setting important precedents in privacy, bias mitigation, and accountability. While not without challenges, state actions are driving a more nuanced and adaptive approach to AI governance, which the federal government has yet to match.

The setup

The rapid advancement of artificial intelligence (AI) technologies has outpaced the development of comprehensive regulatory frameworks. As of October 2023, the US federal government has yet to pass any significant legislation addressing AI’s ethical, legal, and social implications. This inaction has left a regulatory vacuum that state governments are increasingly stepping in to fill. From California’s AI Accountability Act to New York’s Biometric Privacy Law, state legislatures are enacting a patchwork of regulations that aim to address the challenges posed by AI. This trend is gaining momentum as public awareness of AI’s risks—such as algorithmic bias, privacy violations, and job displacement—grows. The conventional wisdom has been that federal regulation is necessary for a coherent national approach. However, the urgency of the situation and the slow pace of federal action have made state-level initiatives a critical component of AI governance.

The argument

1. State laws are more agile and responsive to local needs

One of the key advantages of state-level AI legislation is its ability to be more agile and responsive to local needs and concerns. Unlike federal legislation, which often requires lengthy debates and compromises, state legislatures can act more swiftly to address emerging issues. For example, Illinois passed the Artificial Intelligence Video Interview Act in 2019, which requires employers to notify applicants when AI is used in video interviews and to explain how the AI works. This law was a direct response to the growing use of AI in hiring processes and the concerns about bias and transparency. Similarly, California’s Consumer Privacy Act (CCPA) has been amended to include provisions specific to AI, such as the right to opt-out of automated decision-making. These state laws are often driven by local incidents and public outcry, making them more attuned to the immediate concerns of their constituents.

2. State laws are setting important precedents

State-level AI laws are also setting important precedents that could influence future federal legislation. For instance, New York’s 2021 Biometric Privacy Law, which regulates the collection and use of biometric data, has set a high bar for privacy protection. The law requires companies to obtain explicit consent before collecting biometric data and imposes strict penalties for violations. This has prompted other states, such as Maryland and Washington, to consider similar legislation. Similarly, California’s AI Accountability Act, which mandates transparency and accountability in AI systems, is being watched closely by other states and could serve as a model for federal regulation. These state laws are creating a de facto national standard that companies must adhere to if they want to operate across state lines.

3. State laws are fostering innovation and competition

State-level AI laws are also fostering innovation and competition among tech companies. When states enact stringent regulations, companies are forced to innovate to comply with the new rules. For example, after the implementation of the CCPA, many tech companies developed new tools and processes to ensure compliance, such as automated data mapping and consent management platforms. This has led to the development of a new market for privacy-enhancing technologies. Similarly, the Illinois AI Video Interview Act has spurred the development of new AI systems that are more transparent and less biased. By creating a regulatory environment that encourages responsible AI development, state laws are driving innovation in the tech industry.

4. State laws are highlighting the need for federal action

While state-level AI laws are filling an important gap, they also underscore the need for federal action. The patchwork of state regulations can create confusion and compliance challenges for companies, particularly those operating nationwide. For instance, a company that wants to use AI in hiring processes must navigate the different requirements of each state, which can be costly and time-consuming. This has led to calls for a more coherent national approach to AI regulation. However, the lack of federal action has made state laws a necessary stopgap measure. The success of these laws in addressing AI-related issues could serve as a catalyst for federal legislation, as lawmakers see the benefits of a more unified regulatory framework.

What the conventional view gets wrong

The conventional view that federal legislation is the only viable solution to AI regulation misses several key points. First, it assumes that federal action is inherently superior to state-level initiatives, ignoring the agility and responsiveness of state governments. Second, it overlooks the fact that federal legislation can be slow and cumbersome, often taking years to pass and implement. Third, it fails to recognize the role that state laws can play in setting precedents and influencing federal policy. Finally, it underestimates the ability of state laws to foster innovation and competition by creating a regulatory environment that encourages responsible AI development.

Counter-argument 1: Fragmentation and inconsistency

One of the main criticisms of state-level AI laws is that they lead to fragmentation and inconsistency. Critics argue that a patchwork of regulations can create confusion and compliance challenges for companies, particularly those operating nationwide. For example, a company that wants to use AI in hiring processes must navigate the different requirements of each state, which can be costly and time-consuming. However, this argument ignores the fact that many companies already operate in multiple jurisdictions with different regulatory environments. Moreover, the inconsistency of state laws can be mitigated by the development of industry standards and best practices. For instance, the National Institute of Standards and Technology (NIST) has developed a framework for AI risk management that can help companies navigate the complex regulatory landscape.

Counter-argument 2: Lack of expertise and resources

Another criticism of state-level AI laws is that state governments may lack the expertise and resources to effectively regulate AI. Critics argue that AI is a complex and rapidly evolving field that requires specialized knowledge and technical expertise. While it is true that AI regulation poses significant challenges, state governments have shown a willingness to collaborate with experts and stakeholders to develop effective regulations. For example, California’s AI Accountability Act was developed in collaboration with tech companies, academics, and civil society organizations. Moreover, state governments can leverage the expertise of federal agencies, such as the Federal Trade Commission (FTC) and the National Institute of Standards and Technology (NIST), to inform their regulatory efforts.

The implications

For builders: A call for proactive compliance

For AI developers and companies, the rise of state-level AI laws means that proactive compliance is essential. Companies must stay informed about the regulatory requirements in each state where they operate and adapt their AI systems accordingly. This may involve investing in new technologies and processes to ensure compliance, such as automated data mapping and consent management platforms. Moreover, companies should engage with policymakers and stakeholders to help shape the regulatory environment. By participating in the legislative process, companies can help ensure that regulations are practical and effective.

For users: Greater transparency and accountability

For AI users, state-level AI laws are likely to lead to greater transparency and accountability. As more states enact regulations that require companies to disclose how their AI systems work and to mitigate bias, users will have more information about the AI they interact with. This can help build trust and confidence in AI technologies. Moreover, users will have more recourse if they believe that an AI system has been used unfairly or inappropriately. For example, California’s AI Accountability Act gives users the right to request information about how an AI system works and to file complaints if they believe that the system has been used unfairly.

For regulators: A blueprint for federal action

For federal regulators, state-level AI laws provide a blueprint for future action. By observing the successes and challenges of state-level initiatives, federal policymakers can gain insights into what works and what doesn’t. This can inform the development of a more coherent national approach to AI regulation. Moreover, the success of state laws in addressing AI-related issues could serve as a catalyst for federal legislation, as lawmakers see the benefits of a more unified regulatory framework.

For the industry: A catalyst for innovation

For the AI industry, state-level AI laws are a catalyst for innovation. By creating a regulatory environment that encourages responsible AI development, state laws are driving the development of new technologies and processes. For example, the Illinois AI Video Interview Act has spurred the development of new AI systems that are more transparent and less biased. This has led to the development of a new market for privacy-enhancing technologies. Moreover, the rise of state-level AI laws is likely to lead to increased investment in AI research and development, as companies seek to comply with the new regulations.

What I’m watching next

1. The development of industry standards and best practices

As state-level AI laws continue to proliferate, the development of industry standards and best practices will be crucial. These standards can help ensure that companies comply with the different regulatory requirements and can also serve as a guide for state and federal policymakers. For example, the National Institute of Standards and Technology (NIST) has developed a framework for AI risk management that can help companies navigate the complex regulatory landscape.

2. The role of federal agencies in shaping AI policy

The role of federal agencies, such as the Federal Trade Commission (FTC) and the National Institute of Standards and Technology (NIST), in shaping AI policy will be another key area to watch. These agencies have the expertise and resources to inform the development of AI regulations and can play a crucial role in ensuring that state and federal laws are aligned.

3. The impact of state-level AI laws on AI development and deployment

Finally, the impact of state-level AI laws on AI development and deployment will be important to monitor. As companies adapt to the new regulatory environment, it will be interesting to see how these laws affect the pace and nature of AI innovation. Will they stifle innovation or spur it? The answer will likely depend on the specifics of the regulations and the ability of companies to adapt.

The bottom line

State-level AI laws are filling a critical gap in AI regulation, setting important precedents, and fostering innovation. While not without challenges, these laws are a necessary step toward a more coherent national approach to AI governance. The rise of state-level AI laws underscores the need for federal action, but it also highlights the importance of a collaborative and adaptive approach to AI regulation. As AI technologies continue to evolve, policymakers, companies, and stakeholders must work together to ensure that AI is developed and deployed in a responsible and ethical manner.