Philadelphia is finally starting to explore how to regulate AI

Legislators and governments, including the City of Philadelphia, are just beginning to understand the opportunities and challenges that lay ahead with artificial intelligence.

Councilmember Rue Landau presented a resolution to City Council on Thursday requesting to hold hearings about the state of AI in Philadelphia. While there aren’t any federal laws regulating AI at the moment, Pennsylvania’s state government is trying to catch up, too, with legislation on the docket led by state Sen. Tracy Pennycuick. These proposed regulations also align with concerns about disinformation throughout the current election cycle. 

“We want to get a better understanding [of] what AI technologies are out there, how they’re being used by the private industry and by government,” Landau told Technical.ly. “We want to focus on what the risks are and what the guidelines for it should be.” 

Before laws can be crafted, the council first needs to gather information and assess what the most relevant concerns exist about AI. That’s where the new resolution comes into play.

Landau, chair of the Committee on Technology and Information Services, wants to learn about the implications of AI on the public, private and nonprofit sectors by tapping tech sector experts as well as government entities. 

The purpose of the hearings would be to understand the potential AI has and how it can be used safely, while also understanding Philadelphians’ concerns about the tech, like discrimination and how it will change the workforce. 

The most immediate concern is how AI is contributing to disinformation around the election cycle, Landau said. The resolution specifically cites the fake robocalls that discouraged voters from voting during the New Hampshire Primary earlier this year. 

A majority of Americans are aware of AI, but 52% are more concerned about AI than they are excited, according to Pew Research Center. 

State lawmakers combat deepfakes and AI election fraud

Pennsylvania’s state government is already working to keep up with changing AI technology with a slate of related bills. 

But the state government can only put some regulations on AI to protect people, without going so far that it restricts technology advancement, Pennycuick said. Pennsylvania is behind other states that have already worked to put regulations on AI and deepfakes. 

“We’re so far behind on trying to put guardrails on the usage of artificial intelligence,” Pennycuick told Technical.ly. “Every day there’s something new that comes out on AI.”

Pennycuick’s committee talks to legislators from other states, members of the National Institute for Standards in Technology and other industry experts to keep on top of all the change. They also talk to Big Tech companies like Google and Microsoft as well as AI startups. 

Constituents also reach out detailing what they’re concerned about, that’s partially how legislation about CSAM came about, she said. 

Senate Bill 1213 passed in the Pennsylvania Senate in June to combat deep fakes targeting children. This bill makes sexual deep fakes illegal and clarifies that AI-generated CSAM is illegal. The bill also requires social media platforms to take down images from all sources. 

This seems obvious, but because the technology comes with unprecedented capabilities, previous laws didn’t cover AI.  

Before that, the Pennsylvania House passed an AI content disclosure bill in April that requires people to include a disclosure on any piece of content that was created using AI. 

“We don’t want to see someone using AI to create a product that they didn’t really produce themselves,” she said. 

Pennycuick, along with three other state senators, also introduced Senate Bill 1217 in May, which looks at AI interference in elections. The bill would prohibit the use of generative AI to represent a candidate and possibly influence an election without the candidate’s permission. Pennycuick hopes the bill will pass this fall as election season heats up. 

“We’re trying to strike a balance between free speech and artificially putting words into someone’s mouth,” she said. 

Now, Pennycuick is working on legislation to address right-to-know requests created by data harvesting and AI. Right-to-know requests are a way for people to access public records from government entities. When people use AI to create right-to-know requests for small municipalities for information that is publicly accessible it costs a lot of time and money, she said. 

AI legislation sweeps the states, with little federal guidance

A majority of US states, 45 out of 50, introduced AI legislation this year, according to the National Institute of Standards and Technology. Of those, 31 approved or enacted legislation related to AI.

Like Pennsylvania, for example, states like Alabama, California and North Carolina all have bills in the works against AI-generated CSAM. 

Other legislation around AI varies widely, ranging from guidelines around companies developing the tech to establishing dedicated AI task forces to bills that guide how AI should be integrated into everyday life. 

For example, Tennessee introduced a bill that requires schools to adopt AI use policies for students and staff, but it wasn’t approved. Colorado enacted a bill that requires AI developers to do what they can to avoid discrimination in the technology and provide AI disclosures to consumers. 

There are no federal laws that regulate the development and use of AI at the moment, though there have been guardrails popping up to curb misuse. 

The Biden administration released an executive order last October about safe development of AI. The executive order promoted safe and secure AI tools, responsible innovation and advocacy for equity and civil rights, among other goals. 

In 2022, the Biden administration also crafted an AI Bill of Rights that described five pillars for safe AI development and deployment, including safety, discrimination protections, data privacy, notice of AI being used and the ability to opt out of AI systems. 

Congress is making progress, too. There’s a bill in the Senate that would require federal agencies to use guidelines developed by the National Institute for Standards and Technology on all AI projects. It would provide standard practices and address cybersecurity concerns. 

For now, though, local jurisdictions like Philly move ahead with city council hearings as the first step to figuring out how to start crafting its own approach, Landau said. 

“This is going to be an ongoing process of ours,” Landau said. “We’re really looking forward to embracing what’s out there and figuring out how it’s being used.”