Amazon, Google Among Firms Focusing on AI Lobbying in States

Fear that artificial intelligence could draw unfriendly regulation has led Big Tech to ramp up its lobbying to promote the technology in statehouses nationwide.

Companies such as Amazon.com Inc., Google, and Microsoft Corp. are first focusing on explaining AI technology to legislators from Albany to Austin as lawmakers begin considering additional safeguards and restrictions. The tech giants hope a soft touch now will win goodwill from elected officials before companies begin pushing their policy priorities for less regulation, according to interviews with more than a dozen state lawmakers and industry insiders.

“The first thing is education,” Matt Wing, a political consultant and founder of Wingspan who works with companies like Uber Technologies Inc., said in an interview.

Most lawmakers do “not understand this technology” which presents tech companies with a “valuable” opportunity to define it for them in a more positive way as they begin considering how to regulate it, he said.

The companies are bullish as generative AI—the technology behind tools like OpenAI’s ChatGPT that produce text, images, and video with human-like quality—is powering a market projected by Bloomberg Intelligence to grow from just $40 billion in 2022 to $1.3 trillion in 10 years.

AI companies have demonstrated in recent months the growing importance they place on statehouses, especially as a divided Congress appears unlikely to address the issue heading into a presidential election year.

The lobbying efforts range from an Amazon-sponsored legislative policy panel in Indiana to a Manhattan sit-down between Microsoft and nearly a dozen Democratic legislators. Google officials made sure that one of their executives was among the first to testify before Connecticut lawmakers at the first meeting of the state’s AI task force in September.

“Companies generally are sort of looking for lawmakers who are very thoughtful and smart, maybe have a tech background, who they can really kind of work hand in hand with and build partnerships with to scope narrow and focused regulations,” said Chloe Autio of the Cantellus Group, an advisory firm on tech issues.

‘Build Trust’

AI firms have made normalizing the technology a key part of their education outreach to lawmakers.

The effort includes company officials explaining the difference between large language models and the machine learning that teaches them at events like the National Conference of State Legislatures in Indianapolis in August. Large language models are algorithms that can understand and translate vast amounts of data into readable text and are used by chatbots like ChatGPT, which is backed by Microsoft.

Dispelling fears about the technology is also a key part of such presentations, especially as conspiracy theories have emerged over it use, adding to the unease many people feel about AI.

Google’s Bard, a competitor of ChatGPT, features technology that has more to do with everyday tools that correct spelling and translate foreign languages on the internet than job-killing robots, Beth Tsai, Google’s global policy director for generative AI, said on an Aug. 15 panel.

Tsai said a lack of familiarity with AI could needlessly lead to “knee-jerk fear” of the technology.

“I am here to tell you this is not new and scary,” she said, “because AI has been around for a while and it’s been powering the products that you use every day.”

Safety features prevent Bard from answering “dangerous” questions like how “to build a bomb,” Tsai said in response to a question from one audience member at the legislative conference. “It should not give you that answer,” she added.

Similar messaging about corporate responsibility appeared at an earlier NCSL panel where Nicole Foster of Amazon Web Services spoke about attempts to “build trust” in the technology during a discussion sponsored by the company entitled “Artificial Intelligence: Are We Ready?” Amazon, Google, and Meta Platforms Inc.—which is also making a major push into AI—also exhibited at the summit.

Microsoft has been particularly active in pursuing face time with New York officials; lobbying records show that company representatives have met with at least 10 state lawmakers, as well as the office of Gov. Kathy Hochul (D).

Some of that outreach was related to legislation passed earlier this year that would establish a task force devoted to studying AI and crafting possible policy recommendations.

But the company—the biggest investor in OpenAI—devoted the bulk of such outreach to school Democratic lawmakers on its AI positions at an Aug. 3 meeting in midtown Manhattan.

“They just really wanted to brief us legislators about what their firm is doing and the approach they’re taking with a less-than-subtle context that additional legislation and regulations may not be necessary,” said state Sen. John Liu (D) in an interview.

“Shocking!” Liu said,in jest.

‘It’s Good Politics’

Big Tech’s charm offensive comes as more states pass laws to begin formally studying AI and possible ways of regulating it.

Tsai was among the first to testify at the Sept. 20 inaugural meeting of a Connecticut task force studying artificial intelligence where she highlighted Google’s “view towards responsibility and safety.”

Her expertise in detailing the basic concepts of AI as well as the company’s work with the technology was helpful ahead of possible regulatory action next year, state Sen. James Maroney (D), who is co-chairing the panel, said in an interview.

“The goal is to listen to diverse viewpoints to be able to craft fair legislation and also just to understand from the tech companies what we do to make sure that whatever we pass, they can comply with it,” said Maroney.

Amazon, Google, Microsoft, and other AI developers have detailed their willingness to accept some limits on how AI is used, most recently at a Sept. 14 closed-door hearing organized by Senate Majority Leader Chuck Schumer (D-N.Y.). The Biden administration also has proposed an AI bill of rights that should guide the design, use, and deployment of the technology.

But state-level activity is likely to move at a much faster pace, just like with new privacy laws, and has become a particular focus. “Companies are recognizing that at the federal level, there’s not going to be a lot of progress made,” Autio said.

The companies characterize their outreach to state lawmakers as part of wider efforts to address AI concerns from policymakers while urging them to avoid backing restrictions that would stifle technological development.

“Regulation of high-risk uses of AI can enhance trust in the technology while also providing industry with guidance that eliminates uncertainty and ultimately enhances innovation,” Amazon said in a statement.

Any significant state-level action on AI will likely take place in 2024 when most legislatures are scheduled to convene. Microsoft said in a statement that it aims “to educate, be curious, and learn” in the meantime.

Industry pushback earlier this year stalled a California bill that would have regulated business use of AI technology, showing the strength of the lobbying effort. The sector is a force in American political life, with more than $11.1 million spent by Amazon, Google, and Microsoft alone on state elections since 2018 in California, New York, Texas, Florida, Illinois, and Washington, according to state records.

AI companies are aiming to protect their own profits more than the societal good, and their outreach to state lawmakers must be viewed through that prism, said Shelly Palmer, a Syracuse University media professor.

“Of course, they’re asking to be regulated,” she said. “It’s good politics. They don’t care what the state legislatures do as long as they leave them basically alone.”