Last week’s “Pro-Human Assembly,” attended by both Bernie Sanders and Steve Bannon, highlighted Americans’ broad and bipartisan concern over the risks of artificial intelligence.
Straight Arrow spoke with two invited content creators, who agreed the event felt so bipartisan it bordered on confusing. The emotional center of the event was anchored by speeches from parents whose children had died after chatbot interactions, but both creators suggested that more politicized harms — and the discussion of real solutions — were backburnered in an attempt to appeal to the widest possible group.
Download the Straight Arrow app today to get the stories that matter free from manipulation, bias or agenda.™
Point phone camera here
“In a bipartisan effort, it’s always going to be the kids,” said Clara Fulks, a technologist.
The speeches were “obviously, gut-wrenching,” said AI activist Brian Patrick, “but it wasn’t prescriptive, necessarily, in terms of, ‘This is what we need to do’.”
Their experiences also shed light on a network of Silicon Valley lobbyists who are channeling millions of dollars into shaping AI policy — and on the state-level implications of the federal legislation they support.
READ MORE: Are we all going to die? Inside the push to control AI regulation
What does ‘pro-human’ even mean?
Fulks is both a technologist and content creator. She hadn’t previously worked with the Future of Life Institute, which organized the event.
“I was just familiar with a little bit of their work,” Fulks told Straight Arrow. “So I was like, OK, this feels like a trusted institution; let me go.”
The Institute paid for her travel and hotel, she said, but she was never offered any money for posts or asked to make coverage of the event.
She said the event left her confused, especially when she saw that both Bannon and Sanders had been added as last-minute headline speakers.
“It felt like we’re gathering a lot of very disparate, odd names to gather media attention,” she said.
Fulks told Straight Arrow that she attended Bannon’s talk but walked out in anger because of his association with Jeffrey Epstein.
“This is Epstein’s biggest defender. And you know, 20 minutes ago, we had a panel of parents of children that had been sexually exploited by companion AI chatbots on stage.”
The event was meant to be “Pro-Human,” Fulks said, so she was surprised that those parents were the only example she heard of how the AI industry is currently harming people. She said she would have liked to hear more about the PTSD and insomnia experienced by content moderators in Kenya, the health problems of the people who mine minerals for AI chips or the ways AI can be used by the military.
“There was one person that talked about data centers,” Fulks said. “But nobody talked about climate, nobody talked about other types of child harm.
“You know, nobody talked about digital surveillance.”
When Fulks looked into the event’s funding, she realized that the Future of Life Institute has long-term financial ties to parts of the AI industry.
The Future of Life Institute was initially founded with a $10 million donation from Elon Musk and received just over $1 million from the “effective altruist” organization Coefficient Giving. Another key donor is early Anthropic investor Jaan Tallinn, who is also on the institute’s board of directors. In 2024, Politico reported on this donor network and its campaigning under the headline “AI doomsayers funded by billionaires ramp up lobbying.”
Effective Altruism (EA) is a fairly complex philosophical movement that prioritizes maximizing human well-being and is perhaps best known for its association with imprisoned crypto CEO Sam Bankman-Fried. Its maximization calculations can lead to controversial conclusions, as they can involve making trade-offs between very broad concepts — like, in one thought experiment, optimizing existence for an infinite number of possible future humans, even at the expense of those currently living.
EA organizations like Coefficient Giving have donated millions of dollars to AI research and legislation. Fulks said effective altruism wasn’t mentioned at the Pro-Human Assembly panels she attended.
“In my view, and I still hold this view after attending the assembly, this was a marketing campaign essentially to say, ‘hey, look, we have bipartisan support,’” Fulks said.
“I was shocked at who wasn’t there,” she said. “There were no data workers or people to even talk about data workers.”
‘Informative and less prescriptive’
Patrick is an activist and content creator whose viral videos “posting something INSANE an AI executive said every day in 2026” have received millions of views. He accepted accommodation for the event, but not travel — and similarly wasn’t offered payment to post. He’d been in contact with FLI before and previously discussed collaboration.
“I think there’s a lot of people who have different philosophies and agendas that they’re trying to bring together,” Patrick told Straight Arrow.
He was aware that the FLI has previously been connected to effective altruist organizations, but said that this wasn’t emphasized or mentioned at the event.
“There’s a pro-human declaration and those sorts of things, but it’s unclear to me, you know, where that ultimately lands exactly in terms of policy,” he told Straight Arrow. “I just know that they were trying to put on a very universal event.”
None of the panels that Patrick attended explicitly mentioned any specific proposals around AI regulation or recommendations for the industry, but rather outlined different issues, like the public response to data centers.
Patrick said he had attended a communications panel for creators, where experts recommended “do fun content and don’t micromanage people.” The panel talked about how different AI messaging strategies might land with the general public, but attendees weren’t given explicit instructions on messaging, or offered video templates.
“When I’ve talked to them about potentially working with them on collaborating,” he said, “they haven’t said anything about how you need to do X, Y, and Z.”
‘My heart sank’
The only panel at the event that Fulks said focused on specific policy was a talk by U.S. Rep. Lori Trahan, D-Mass., one of two sponsors of a federal bill to regulate AI known as the FRONTIER Act.
“My heart sank,” Fulks said, when she saw Trahan on the speaker list. “Because that bill is just so predatory toward — especially — California AI legislation”.
Trahan and U.S. Rep. Jay Obernolte, R-Calif., introduced the bipartisan bill in July, and it includes a provision that would ban states from creating legislation that conflicts with it. Last week, OpenAI CEO Sam Altman publicly endorsed the measure.
Since there is no federal legislation on AI, state laws are currently the main source of regulation.
The major AI companies are broadly opposed to state-level laws, and the battle around them has been playing out since last winter, when President Donald Trump signed an executive order to block states from enacting AI regulation.
But Politico recently described a “defiant mood” sweeping state governments, suggesting that individual legislatures are shedding “their fear of Silicon Valley,” ignoring industry lobbying groups and moving to create regulation against specific technological harms within their jurisdictions.
California has been among the busiest of these, which is why Fulks sees the FRONTIER Act as specifically targeting the state. Gov. Gavin Newsom responded to the ongoing AI safety debate with an executive order last Friday, calling for stricter regulation and the implementation of a “kill switch” for models.
“We’re not waiting to act,” Newsom said in an announcement. “While Washington abdicates its responsibility to protect Americans, California is building on the strongest AI regulatory framework in the nation.”
READ MORE: How AI’s threats are uniting Bernie Sanders and Steve Bannon in a populist crusade against Big Tech
Teeth — but not super sharp
Any good federal law will tend to block states from making rules that conflict with it, said Stephen Casper, a computer scientist and assistant professor of public policy at Harvard University.
“Usually when people are upset about preemption, it might partially have something to do with the state autonomy argument, but usually it’s because there is a bill that is attaching preemption to something that is not expansive enough,” he told Straight Arrow.
He sees the act as “the most serious proposal on the table at the federal level for doing something pretty meaningful,” but said there are several things he would do to improve it.
The FRONTIER Act is an evolution of a prior bill called the Great American AI Act. That first draft was deservedly criticized, Casper said, because the scope of the state-level ban was greater than the federal rules that would have been created to replace them.
“It would have preempted state laws on child safety without having anything inside of the bill itself to address those same problems.”
Those issues don’t exist in the new version, he said.
However, the revised bill would still remove states’ ability to create regulation that is more stringent or rigorous than the federal level. For example, a current Massachusetts bill, which Casper thinks is likely to pass, would be much stricter than the requirements for the FRONTIER Act.
“It requires large frontier model developers to publish a risk report and update it every six months,” Casper said. “It’s not just a single-model analysis. It’s not a safety framework. It’s a company-wide risk report,” he added. “This is, to my knowledge, unique among state laws.
“And when I say ‘to my knowledge’, I mean I’m confident,” he said. “I read all the state laws.”
If the FRONTIER Act is passed, he said, states would be unable to ask for additional documentation or regulation past the federal level. “It would be illegal,” he said.
“The FRONTIER Act, does leave a lot to companies,” he said. “But it at least requires some transparency about these things and allows companies to be held accountable to some requirements and some consequences.”
The bill would still give AI companies a lot of freedom to choose their own risk management frameworks and protocols, he said, but “they have to be open about those things.”
Casper would like to see more criminal penalties for destroying records and instances of executives lying about the dangers of AI products, similar to how the financial services industry works.
Round out your reading
- Lawyer says 57 Cornell frat members were in the group chat. Could their silence be a crime?
- Florida county discovers mysterious Flock cameras with no obvious owner.
- Off Script: They won their primaries on one message. Now it’s disappearing.
- Some question whether a popular affordable housing policy is making the nation’s home shortage worse.
- Kelvin wave incoming: Why the US should care.
