AI and consumer rights

A few things are happening at once. Virtually every company is transitioning into an artificial intelligence company and is actively assessing not only how to use AI tools but also how to build and develop their own. 

At the same time, U.S. states are passing more comprehensive privacy laws that are creating affirmative rights for consumers in relation to their data, which has potential implications for AI developers. 

Legislators, primarily at the state level, are specifically considering and, in some cases, passing new laws focused on generative AI. There is also now an influx of tools that allow consumers to more easily exercise their rights under these new laws. 

All the while, regulators are paying particular attention to whether companies are honoring consumer rights, and the — albeit limited — enforcement actions brought under these new laws primarily focus on companies' failures to comply with consumer rights requirements.  

The confluence of these factors raises a series of questions for both companies and consumers: What rights do consumers have in relation to their data being used to train AI models? Does this extend to all their data? Conversely, what obligations do companies have to honor such requests? What is the business rationale for offering consumers broad-based rights in relation to AI training, regardless of what a company's legal obligations are? How will the answers to these questions change going forward?

Privacy law and AI development

Existing privacy law has answers to some of these questions already. To the extent a company relies on "personal information" or "personal data" to train its models, it must evaluate — at least in the U.S. — how the more than 20 state comprehensive privacy laws impact what data it can use and what obligations it has to the individuals to whom the data relates.