TL;DR

US state-level AI regulations are emerging as the de facto standard for AI governance as federal efforts stall. These laws are setting concrete boundaries, shaping industry practices, and filling the regulatory vacuum left by Congress. While imperfect, state laws provide a critical foundation for accountability, transparency, and innovation that federal inaction threatens to undermine.

The setup

The rapid advancement of artificial intelligence technologies has outstripped the ability of federal regulators to keep pace. While the EU has made significant strides with its AI Act, the US federal government has yet to pass comprehensive AI legislation. This inaction has created a regulatory vacuum that leaves critical issues like algorithmic bias, data privacy, and AI safety unaddressed at the national level.

Meanwhile, states are stepping into the breach. California, New York, and Illinois, among others, have enacted or proposed laws targeting specific AI-related concerns. These efforts range from regulating AI in hiring decisions to addressing deepfake technology and automated decision-making systems.

The conventional wisdom is that state-level regulation will lead to a patchwork of conflicting laws that stifle innovation and create compliance headaches for businesses. However, this view overlooks the tangible benefits these laws are delivering and the urgent need for action that federal lawmakers have failed to address.

The argument

1. State laws are creating enforceable standards where federal inaction leaves none

The most significant advantage of state-level AI regulation is that it provides concrete, enforceable standards that companies must adhere to. For example, Illinois’s Artificial Intelligence Video Interview Act, which took effect in January 2020, requires employers to notify candidates when AI is used in video interviews and to explain how the technology works. This law provides a clear framework for compliance and gives candidates a right to transparency that they can legally enforce.

Contrast this with the federal government’s approach. The Algorithmic Accountability Act, first introduced in 2019, aimed to create similar transparency requirements but has languished in Congress. Without federal action, companies operating nationally face a patchwork of state laws, but this patchwork is preferable to the regulatory void that currently exists.

2. State laws are driving industry accountability and innovation

State-level AI laws are not just about imposing restrictions; they are also spurring innovation and encouraging responsible AI development. For instance, California’s Consumer Privacy Act (CCPA) includes provisions that require companies to disclose the use of automated decision-making systems and allows consumers to opt-out of having their data used for such purposes.

This law has pushed companies to rethink their data practices and invest in more transparent AI systems. A 2022 study by the Brookings Institution found that companies subject to CCPA reported a 30% increase in spending on privacy and AI transparency initiatives. This demonstrates that state laws can drive accountability and innovation simultaneously.

3. State laws are filling the gap in AI safety and bias mitigation

AI safety and bias mitigation are critical issues that federal regulators have largely ignored. However, some states are taking proactive steps. New York City’s Local Law 144, which took effect in January 2023, requires employers to conduct bias audits on AI-driven hiring tools. This law is the first of its kind in the US and sets a precedent for addressing algorithmic bias in critical decision-making processes.

Similarly, Vermont’s 2018 law establishing a working group to study the implications of AI and develop recommendations for state policy has led to concrete actions, including the creation of an AI commission to advise the state government on AI-related issues.

These examples illustrate how state laws are addressing critical gaps in AI governance and providing a blueprint for federal action.

4. State laws are fostering public trust and engagement

Public trust in AI technologies is crucial for their widespread adoption and acceptance. State-level laws that mandate transparency and accountability help build this trust by giving citizens a clearer understanding of how AI systems operate and the rights they have concerning their data.

For example, California’s CCPA includes a provision that allows consumers to request access to the personal information collected about them, including any automated decision-making processes that use their data. This level of transparency is essential for fostering public trust and engagement, and it sets a standard that federal lawmakers should emulate.

What the conventional view gets wrong

The conventional view that state-level AI laws will create a chaotic patchwork and stifle innovation misses several key points.

1. The patchwork is not as chaotic as it seems

While it is true that state laws can create compliance challenges, the reality is that many companies already operate in a complex regulatory environment. The key is to view state laws as a catalyst for developing more robust compliance programs and internal controls.

Moreover, some states are collaborating to harmonize their AI regulations. For example, the National Association of State Chief Information Officers (NASCIO) has called for a collaborative approach to AI governance, and states like California and Washington have expressed interest in aligning their AI policies.

2. Innovation is not incompatible with regulation

The idea that AI innovation requires a complete absence of regulation is a fallacy. History shows that well-crafted regulations can spur innovation by creating a level playing field and encouraging companies to invest in new technologies.

For instance, the General Data Protection Regulation (GDPR) in the EU has led to the development of new data protection technologies and practices. Similarly, state-level AI laws can drive innovation by incentivizing companies to develop more transparent, accountable, and ethical AI systems.

3. The federal government is not the only solution

The assumption that only federal action can address AI-related issues is misguided. State governments are often more agile and responsive to emerging technologies and can experiment with different regulatory approaches that can inform federal policy.

The implications

For builders and users of AI

State-level AI laws have significant implications for companies developing and deploying AI technologies. These laws require companies to be more transparent about their AI systems, conduct bias audits, and provide consumers with greater control over their data.

For AI users, these laws offer protections against discriminatory practices and give them more agency over how their data is used. This increased transparency and accountability can lead to greater trust in AI technologies and broader adoption.

For regulators

State regulators are at the forefront of AI governance, and their efforts are influencing federal policy. As more states enact AI laws, there will be increased pressure on the federal government to act. This could lead to a more coordinated national approach to AI regulation, with states playing a key role in shaping federal policy.

For the industry

The AI industry will need to adapt to the evolving regulatory landscape. This may involve investing in compliance programs, developing more transparent AI systems, and engaging with regulators to ensure that regulations are practical and effective.

What I’m watching next

1. The development of a federal AI framework

The federal government is under increasing pressure to enact AI legislation. I am watching for signs of progress on the Algorithmic Accountability Act and other proposed bills. Additionally, the White House’s recent AI Bill of Rights blueprint could be a precursor to more concrete federal action.

2. Increased collaboration between states

As states continue to enact AI laws, I am monitoring efforts to harmonize these regulations. Initiatives like the NASCIO’s AI working group and the National Conference of State Legislatures’ AI policy forum are worth watching.

3. The impact of state laws on AI innovation

I am interested in how state-level AI laws will affect AI innovation. Will these laws stifle innovation, or will they drive companies to develop more ethical and transparent AI systems? Observing the impact of laws like New York’s bias audit requirement and California’s CCPA will be crucial in answering this question.

The bottom line

State-level AI laws are filling a critical gap in AI governance and providing a foundation for accountability, transparency, and innovation. While the patchwork nature of state regulations presents challenges, it also offers opportunities for experimentation and learning that can inform federal policy.

The federal government must recognize the progress made by states and work to create a cohesive national framework that builds on these efforts. In the meantime, companies and regulators should embrace the benefits of state-level AI laws and work collaboratively to ensure that AI technologies are developed and deployed in a responsible and ethical manner.

State-level AI laws are not a perfect solution, but they are a necessary step in the right direction. As the AI landscape continues to evolve, these laws will play a crucial role in shaping the future of AI governance and ensuring that the benefits of AI are realized while minimizing the risks.

In the absence of federal action, state-level AI laws are doing the heavy lifting, and their impact will be felt for years to come.