AI’s next big blue battleground
There’s a lot going on in Illinois
A few weeks back, I got a tip: accelerationist AI lobbyists and industry representatives were said to be in talks with state assembly leadership in Illinois to bend pending AI legislation making its way through the legislature towards their ‘light-touch’ preferences. It sounded like the replay of battles that happened in California and NY: in each state, lawmakers had drawn up stringent AI regulations only for them to be weakened (and in the case of California, rivaled by a more light-touch bill) in negotiations heavily influenced by industry.
I wasn’t able to confirm that exact tip — my reporting did not reveal a coordinated effort among legislative leaders or at governor level. But it did lead me to dive into the complexity of attempts to regulate AI in Illinois, which are being heavily influenced by both industry representatives and safety advocates, not to mention millions of dollars in political spending. The result is a flurry of AI bills, with varying degrees of industry-friendliness, that make the Land of Lincoln a new battleground for AI regulation.
“After California and New York took the first steps last year to manage risks from AI, legislators in Illinois tell me they are ready to take the next steps to ensure their kids and communities are protected,” said Scott Wisor, Policy Director at the Secure AI Project. “As the country’s fifth largest state economy, I anticipate all eyes will now be on how the legislature and Governor Pritzker advance AI safety policy, as ~90% of their constituents say they want.”
AI regulation isn’t a brand-new topic for Illinois lawmakers. The state passed amendments to the Illinois Human Rights Act in 2024 to prohibit AI-based workplace discrimination, which went into effect on January 1. It also passed a bill requiring employers to notify job applicants when AI is used in video interviews all the way back in 2019.
But this year, the introduction of AI-related bills has exploded. Many are so-called “messaging bills,” meaning they likely won’t move out of committee but signal the lawmakers’ intentions to voters, but a few appear to have legs. State Representative Daniel Didech, for example, has introduced HB 4705, which is similar to California’s SB53 and New York’s RAISE Act, but with additional child safety components, such as a mechanism for reporting child safety risks and implementing and publishing a child safety plan audited by third parties. That bill is supported by the Secure AI project.
State Senator Rachel Ventura, meanwhile, has introduced a large package of AI bills, two of which she calls “heavy hitters.” One, SB 3890, is an expansive data privacy bill, giving citizens the option to opt out of certain types of data collection and use, and requiring companies to be more transparent about how they use and collect consumer data. The second, SB 3502, opens AI developers up to class action lawsuits and is deliberately expansive in order to force slower and more deliberate AI development.
“I think that fear — a little bit — or the caution of ‘we don’t want to get sued,’ will maybe encourage these companies to do a little bit more research or a little bit more trial and error before they put products out there,” Senator Ventura told Transformer.
Ventura’s office said that they have shared their bills with Apple, Google and industry association TechNet for feedback. However, over the phone, Ventura said she hadn’t received any meaningful response, though her office followed-up to say they had been attempting to schedule meetings with lobbyists representing relevant companies.
Ninia Linero, TechNet’s Executive Director for Illinois and the Midwest, told Transformer in a statement, “there are a number of AI bills under consideration in the state, and we are committed to working with Sen. Ventura and Illinois Senate leadership to ensure a thoughtful policymaking process. Illinois leads across many sectors, including technology, and we look forward to helping preserve an environment that supports continued innovation in the state.”
No industry advocates had testified or otherwise publicly commented specifically on Ventura or Didech’s bills, but their preferences are being expressed in other ways. For example, SB 3444, sponsored by state Senator Bill Cunningham, is practically the inverse of SB 3502, protecting frontier AI developers from liability in cases where they followed a set of light-touch safety protocols, very similar to an idea floated by OpenAI’s chief global affairs officer Chris Lehane. Cunningham is a leader in the Illinois Senate, holding the role of president pro tempore. Senator Cunningham’s office and OpenAI did not respond to a request for comment.
Separately, the industry is an active player in campaign fundraising in the state. Meta, via a PAC called Making Our Tomorrow, has spent more than $560,000 on state races according to the Illinois State Board of Elections. The PAC spent on four candidates — Paul Kendrick, Adam Braun, Aja Kearney, and Jaime Andrade — on the basis of curbing legislation that threatens their AI investments.
Advocates who spoke with Transformer said the political spending is less about each candidate’s specific AI policy (none have led initiatives particularly friendly or unfavorable to AI companies) but rather to influence their key votes in the senate on broad statewide initiatives led by governor JB Pritzker, such as a two year moratorium on tax incentives for data center development and several social media policies. Of those Meta supported, only Kendrick won.
There’s also a significant sum in campaign contributions coming from employeesof AI firms. Anthropic employees, who are generally aligned with AI safety groups, have contributed $11,200. Of that $5,500 went to Didech’s campaign from Daniel Ziegler, the senior manager who is also the sole funder of a New York state PAC backing AI safety candidate Alex Bores. Another $4,500 of money given to Didech’s campaign, the reelection campaign for state representative Laura Faver Dias, and the reelection campaign for state representative Jennifer Gong-Gershowitz was from Steven Bills, who has also been politically active in other states according to our elections tracker. Those listing Google on their filings, meanwhile, have given more than $137,000. The largest gift came from former Google CEO Eric Schmidt, who gave $50,000 to former White House chief of staff Rahm Emanuel’s campaign committee for Chicago mayor. Schmidt left Google in 2020 but retains significant holdings in parent company Alphabet.
At the federal level, of course, there’s even more AI money involved. Leading the Future and venture investor Ron Conway fund a Democratic super PAC that contributed a combined $2.52m to the campaigns of Jesse Jackson Jr. and Melissa Bean for House seats in Illinois. Leading the Future is itself a super PAC with funding from Ben Horowitz and Marc Andreessen of venture firm a16z, Perplexity, and OpenAI President Greg Brockman and his wife Anna. Rival AI Safety super PAC Public First Action — which counts Anthropic as its only disclosed donor — initially made a filing indicating they would spend $1m opposing Jackson as well. However, the PAC reversed course after Illinois Democrats and the Congressional Black Caucus expressed discontent over the funding being disclosed on a day Jackson was attending a memorial for his recently deceased father. Jackson lost his election, while Bean, who also received a significant boost from a pro-Israel super PAC, won her race.
Those millions of dollars were distributed in order to influence tech policy up and down the ballot — but the industry’s mixed success suggests that voters in Illinois may actually care enough about AI to make simply throwing money at candidate campaigns less effective than it has been elsewhere. In the state races in particular, opposition candidates drew attention to tech industry political spending, with some candidates even attempting to publicly distance themselves from tech company donations. Illinois’ federal candidates’ embrace of AI in their campaigns had mixed results, too, with Jesse Jackson Jr’s use of an AI-generated voiceover in an ad generating unease.
Marjorie Connolly, communications director at the Tech Oversight Project, argues the failure of some industry-backed candidates to come out on top has two implications for the role of money in other races. Politicians might be less eager to accept AI money, and, if they do, the funding might also prompt attacks on a campaign for being too close to corporations — something voters care about, at least in Illinois. Accepting AI money will “embolden accountability advocates, and make candidates think twice about accepting this support,” she says.
That doesn’t mean that there’s a large cohort of single-issue AI voters in Illinois, but the issue is growing in salience. According to February polling from Impact Research shared with Transformer, which was commissioned by the Secure AI Project and Encode AI, 59% of Democrats and 56% of Independents in the deep blue state want to see more regulations on major technology companies, while 85% of that same group wanted to see legislation regulating catastrophic risks from AI. Nine in 10 of those voters said they were against legislation that exempted AI companies from legal liability. Politicians are seizing upon those voter sentiments to make AI policy a larger issue in the state.
Excellent article! Very informative, especially for those living in Illinois.