In September, California signed into law a series of AI-related measures. What do these laws mean for you and your business?

What Was Signed Into Law?

In total, California passed 18 laws, all consequential in their way. The laws cover a range of AI areas:

  • AI Development Process. AB-2013 requires transparency in training data, the data from which an AI system learns patterns.
  • Risk Management: Government agencies have new requirements to assess the risk of AI systems in catastrophic scenarios (SB-896).
  • Healthcare: The new laws provide patient protection when Generative AI is used by healthcare providers, including instructions for disclosure and proper oversight of AI tools used in patient care (AB-3030 and SB-1120).
  • Privacy: The new laws extend the state's privacy protections to include AI-exposure user information (AB-1008). Further personal protections (mainly driven by the entertainment industry) prohibit generative AI content of a person's likeness without permission (AB-2602 and AB-1836).
  • Legal Definition of AI: The definition provided by California for AI is: "an engineered or machine-based system that varies in its level of autonomy and that can, for explicit or implicit objectives, infer from the input it receives how to generate outputs that can influence physical or virtual environments" (AB-2885).
  • AI Literacy: The new laws require the California Board of Education to consider AI Literacy in math, science, etc. (AB-2876). A further regulation requires superintendents to explore challenges and opportunities for AI in public education (SB-1288).
  • Deepfakes: AB-2905 requires that robocalls disclose when deep fake-created voices are used. Meanwhile, AB-1831, SB-926, and SB-891 strive to protect individuals from exploitation via deepfakes. To prevent deepfakes from influencing election content, AB-2655, AB-2839, and AB-2355 address a range of control measures, from platform labeling to removal and disclosure by creators.

What Did Not Pass Into Law?

SB-1047, a bill seeking to regulate large AI systems, was vetoed. This bill would have required AI developers to document AI creation practices and develop safety and security protocols, as well as provide whistleblower protections for employees at AI companies.

What Do These Mean For You and Your Business?

These laws are wide-ranging, covering everything from personal protection to legal definitions. The laws passed will impact all aspects of AI moving forward. And given our increasingly global landscape, businesses will likely want to pay attention to these laws even if their primary business is not in California. These are also likely the first of many AI laws that we will see coming out in the next few years. Even the vetoed law represents a public desire to see greater regulation of AI development, and similar laws are likely to appear again.

For a business, the key takeaways are:

  • Keep up with the arrival of new laws and also with the first case law in each area. Such laws are often open to interpretation, and the first cases provide great insight into how the legal system will choose to apply them. This will need to be a continual process to track not just these laws but all that will follow.
  • If your company uses AI (as virtually all now do), it would be wise to have guidelines for your employees on how they can remain compliant while using AI to maximize your business's effectiveness.

Source: Forbes

View the original article here.