Jon Potter is a Associate and leads the State-Degree AI Follow at RXN Group. He’s an skilled lawyer, lobbyist, and communicator, has based and led two trade associations and a client group, and consulted many industries and organizations on legislative, communications and problem advocacy challenges. Jon based the Digital Media Affiliation, Fan Freedom, and the Utility Builders Alliance, was Government Vice President of the worldwide communications agency Burson-Marsteller, and a lawyer for a number of years with the agency of Weil, Gotshal, & Manges.
As each a consumer and marketing consultant, Jon has overseen federal, multistate and worldwide advocacy campaigns and engaged lobbyists, communications corporations, and legislation corporations on three continents. Jon has testified in Congress and state legislatures a number of instances, has spoken at dozens of conferences all through the U.S. and internationally, and has been interviewed on nationwide and native radio and tv information applications, together with CNN, In the present day Present, and 60 Minutes.
Are you able to present an summary of the important thing developments in AI laws throughout states in 2024?
2024 has been a rare 12 months for state-level AI laws, marked by a number of developments.
The primary pattern is quantity. 445 AI payments had been launched throughout 40 states, and we count on this may proceed in 2025.
A second pattern is a constant dichotomy—payments about authorities use of AI had been usually optimistic, whereas payments about AI usually and personal sector use of AI had been skeptical and fearful. Moreover, a number of states handed payments creating AI “job forces,” which are actually assembly.
What are the primary considerations driving state legislators to introduce AI payments, and the way do these considerations fluctuate from state to state?
Many legislators need authorities companies to enhance with AI – to ship higher providers extra effectively.
Amongst skeptics, subjects of concern embrace fraudulent and abusive “deepfakes” associated to elections, artistic arts, and bullying; algorithmic discrimination; worry of AI-influenced “life important” selections and determination processes; private privateness and private knowledge use; and job displacement. Some considerations could be addressed in very particular laws, akin to Tennessee’s ELVIS Act and California’s political deepfakes prohibition. Different considerations, akin to dangers of algorithmic discrimination and job displacement, are amorphous and so the legislative proposals are broad, non-specific, and of nice concern.
Some lawmakers imagine that at this time’s social media and digital privateness challenges may have been mitigated by prophylactic laws, so they’re speeding to go legal guidelines to resolve AI issues they worry will develop. After all, it’s very laborious to outline clear compliance steering earlier than precise issues emerge.
How can states craft AI laws that encourages innovation whereas additionally addressing potential dangers?
Laws can do each when it regulates particular use instances and dangers however not foundational multipurpose know-how. An excellent instance of that is the federal legal guidelines that govern makes use of of well being, monetary, and scholar training knowledge, however don’t regulate computer systems, servers, or cloud computing. By not regulating multipurpose instruments akin to knowledge storage and knowledge processing applied sciences (together with AI), the legal guidelines tackle actual dangers and outline clear compliance guidelines.
It’s vital that legislators hear from a variety of stakeholders earlier than passing new legal guidelines. Headlines recommend that AI is dominated by large corporations investing billions of {dollars} to construct terribly highly effective and threat fashions. However there are literally thousands of small and native corporations utilizing AI to construct recycling, office bias, small enterprise lending, and cybersecurity options. Authorized Help organizations and native nonprofits are utilizing AI to assist underserved communities. Lawmakers should be assured that AI-skeptical laws doesn’t shut down small, native, and public-benefit AI exercise.
Out of your expertise, what are probably the most vital impacts that latest AI payments have had on companies? Are there particular industries which were extra affected than others?
We don’t but know the impacts of latest AI payments as a result of only a few latest payments have grow to be legislation and the brand new legal guidelines aren’t but making a distinction. The broadest legislation, in Colorado, doesn’t take impact till 2026 and most stakeholders, together with the Governor and sponsor, anticipate vital amendments earlier than the efficient date. The ELVIS Act in Tennessee and the California deepfakes legal guidelines ought to cut back fraudulent and felony exercise, and hopefully gained’t inhibit parody or different protected speech.
With so many states stepping as much as legislate AI, how do you see the connection between state-level AI laws and potential federal motion evolving?
This can be a transferring goal with many variables. There are already a number of areas of legislation the place the federal and state governments co-exist, and AI is related to lots of them. For instance, there are federal and state office and monetary providers discrimination legal guidelines which might be efficient no matter whether or not AI is utilized by alleged unhealthy actors. One query for legislators is why new laws or laws are wanted solely as a result of AI is utilized in an exercise.
What are the widespread pitfalls or challenges that state legislators face when drafting AI-related payments? How can these be prevented?
It’s all about training – taking the time to seek the advice of many stakeholders and perceive how AI works in the true world. Laws based mostly on worry of the unknown won’t ever be balanced, and can at all times inhibit innovation and AI for good. Different nations will fill the innovation void if the US cedes our management on account of worry.
Are you able to share examples of profitable advocacy that influenced AI laws in favor of innovation?
In Colorado, the Rocky Mountain AI Curiosity Group and AI Salon rallied builders and startups to have interaction with legislators for the primary time. With out lobbyists or insider consultants, these teams tapped right into a wellspring of sensible, sad, and motivated founders who expressed their displeasure in crisp, efficient, testimony and the media – and had been heard.
Equally in California, founders of small AI-forward corporations testified passionately within the legislature and linked with the media to specific pressing concern and disappointment about well-intended however terribly overbroad laws. LIke their counterparts in Colorado, these founders had been non-traditional, extremely motivated, and really efficient.
How do you establish and interact stakeholders successfully in state-level AI coverage battles?
Partly by speaking to individuals such as you, and spreading the phrase that laws has or will quickly be launched and will affect somebody’s livelihood, enterprise, or alternative. Legislators solely know what they know, and what they be taught by speaking to lobbyists and firms they’re conscious of. It’s vital to have interaction within the course of when payments are drafted and amended, as a result of should you’re not on the desk then you definately’re most likely on the menu. Each firm and group that’s constructing or utilizing AI ought to take part earlier than the legal guidelines and laws are written, as a result of after they’re written it’s steadily too late.
What are the best methods to speak the complexities of AI to state legislators?
Advocacy exists in lots of kinds. Whether or not it’s assembly with legislators in-person or by video, sending a letter or electronic mail, or talking with the media – every of those is a approach to make your voice heard. What’s most vital is readability and ease, and telling your individual story which is what finest. Totally different state legislatures have totally different guidelines, processes, and norms, however virtually all legislators are desirous to be taught and wish to hear from constituents.
Thanks for the good interview, readers who want to be taught extra ought to go to RXN Group.
