Lawmakers Respond to AI-Induced Job Displacement
The flip side of AI technology’s economy-boosting potential is the assumption that if AI succeeds as currently imagined, it’ll replace much of today’s workforce. Unsurprisingly, this is a major concern for state lawmakers and their constituents. We’re already seeing headlines of companies using generative AI to replace jobs (e.g., Sports Illustrated and Duolingo) and while the recent New York Times’ lawsuit against OpenAI is directly about copyright infringement, it’s also a response to the existential threat that AI represents to the journalism industry. Policymakers are again placed in the difficult position of balancing protections for individual workforces against the economic productivity gains promised by AI. So far, state lawmakers have proposed several strategies to address this issue.
Last year, New York Assemblymember Pat Burke (D) took a broad approach by introducing legislation (NY AB 8179) that would tax companies for every employee displaced due to automation, which could include AI, manufacturing robots, and other technologies. The amount of the tax paid by the employer in the proposed bill would be the equivalent of any taxes or fees that the state or local government would have collected on that replaced employee’s wages. This is an aggressive approach; unsurprisingly, the bill did not make it out of committee last year.
Legislation like this arises from the very real fear that new technology will replace human workers, causing massive unemployment. It’s a concern as old as capitalism itself. Economists say it’s more complicated than that. They argue that advances in technology will boost production over time, growing the economic pie for the overall benefit of everyone. But that doesn’t mean that individual jobs won’t be replaced. When was the last time you needed an elevator operator, lamplighter, or telephone switchboard operator?
Proponents of AI argue that in the short term, AI will complement much of today’s modern workforce instead of replacing us outright, helping workers become more productive. They point to a history of fear-mongering about economic ruin every time a transformative technology comes along, yet our economic fortunes have not only survived these past technological waves but have prospered. But is today’s latest technological transformation faster, smarter, and more widely applicable than yesterday’s automated elevators, electric lights, and automated telephone switchboard systems? Is AI different? State lawmakers aren’t waiting to find out.
This year’s iteration of AI worker protection bills has taken a more industry-specific approach. This week, Tennessee Governor Bill Lee (R) announced plans to introduce the Ensuring Likeness Voice and Image Security Act (ELVIS Act), which would build upon existing state rules protecting against the unauthorized use of someone’s likeness by adding “voice” to the state’s Protection of Personal Rights law, including civil and criminal penalties for violators. Governor Lee told reporters, “This is really a piece of legislation that will protect voices from technology from artificial intelligence that actually rob and steal an artist's gift. We hope it'll be a blueprint for the country.
California has focused on Hollywood, especially after generative AI was a major factor in last summer’s screenwriter strike. Assemblymember Ash Kalra (D) hopes to move her bill (CA AB 459), which was amended in the waning days of the 2023 legislative session to allow performers to nullify contract provisions that could replace them with a digital replica of their voice or likeness or train generative AI to do so, applying retroactively. Similarly targeting the film industry’s potential use of generative AI to replace human performers is a New York bill (NY AB 7634) that would bar a production company using AI for that purpose from applying for the Empire State Film Production Tax Credit.
A less direct approach is for states to fund job training and career development programs for workers whose jobs have been lost or are endangered by automation (NJ SB 1363) or to produce impact reports on AI’s effects on the job market (NY AB 7838). We’re still early in this process, with more speculation than real-world examples of AI replacing workers, but policymakers will continue to grapple with these challenges and develop new ways to protect jobs from AI. This is not an issue that’s going away anytime soon.
Recent Policy Developments
State-Level Action on Deepfakes: Since the new year, 13 states have already introduced or prefiled bills seeking to regulate deepfakes, looking to join the 5 states that have enacted legislation on electoral deepfakes and the 9 states that have enacted legislation directly targeting sexual deepfakes. With nearly two-thirds of states yet to convene, expect to see more states consider deepfake legislation in the coming weeks and months.
Maryland: Governor Wes Moore (D) issued an executive order on Jan. 8, 2024, establishing principles for state agencies use of AI and establishing an “AI Subcabinet” to promote the AI principles established in the executive order, provide advice and recommendations to the governor on AI matters, and facilitate the statewide coordination on the responsible use of AI.
Nebraska: Sen. John Cavanaugh plans to introduce legislation regulating deepfakes in political advertising similar to a law passed in Michigan last year. Some Nebraska lawmakers recently had their voices cloned as part of a demonstration of deepfake audio, with Sen. Jana Hughes calling the cloned voice “scary close, and it’s only going to get better.”
New Jersey: Governor Phil Murphy (D) called for an “AI moonshot” in his State of the State address on Tuesday, pledging his state would “pioneer breakthroughs” and “harness generative AI to unlock a new generation of possibilities.” Last December, Murphy announced a partnership with Princeton University to develop an AI innovation hub.
New York: In her State of the State address on Tuesday, Governor Kathy Hochul (D) proposed directing $275 million in state funds towards a research center dedicated to artificial intelligence. Working with New York universities, the center would provide the computing power necessary to run modern AI software to provide to partners.