No, AI Companies Shouldn’t Regulate Themselves

AI self-regulation is not an option

This week, a former Anthropic and OpenAI researcher warned on social media that “neither company is acting responsibly” and that they’re “gambling with our lives.” His statement joins a growing list of concerns about AI’s impact on society—some more sci-fi-esque than others—and makes one thing evident: The time to regulate AI is now, and the public must have a place at the decision-making table. 

In a new blog post, Roosevelt President and CEO Elizabeth Wilkins argues that we can’t leave AI safety to private companies alone.

“Yes, this conversation is about risk. But more than anything, it’s about power and control,” Wilkins writes. “Who decides what risks we take: us or them.”

Movement on policymaking at the state level, along with widespread anti–data center protests, clearly show an appetite for action and desire for democratic control over AI technologies.

“We don’t need to settle the debate over exactly how dangerous AI is, or prove our regulators are fully up to the task, before acting,” Wilkins writes.

Read the blog post: Fear Itself: AI Anxiety Is No Excuse for Self-Regulation

More tax brackets, fewer problems

Inequality is through the roof, the top 10 billionaires in the world all live in the US, and rising debt servicing costs threaten the entire economy’s stability. Yet, our top tax rate is only 37 percent—far lower than the 70 percent average during most of the 20th century—and it applies equally to a neurosurgeon making $800,000 and a Fortune 500 CEO amassing tens of millions in income a year. 

In a new brief covered by Politico Weekly Tax, Roosevelt Fellow Samarth Gupta proposes one solution: set up additional tax brackets.

This would allow the government and policymakers to tax extremely high earners significantly more, collect more revenue for the public investments we need, and more easily adjust tax rates based on changing economic conditions and fiscal needs.

Read the brief: Reaganism Broke US Tax Brackets. It’s Time To Fix Them.

Reducing the mind-boggling level of inequality we see today is a critical step toward building a truly democratic society. Join Roosevelt today at 2:00 pm ET for a conversation with Gupta, Senior Fellow Brian Galle, and political scientist Nate Kelly about using the tax code to shift economic and political power away from a small group of the ultra-wealthy and toward the entire working class. Register here. 

What else we’re up to

• When it comes to Social Security reform, starting with the payroll tax cap is a no- brainer. A small portion of workers earn more than the maximum amount of income taxed as Social Security contributions, but the pool of untapped revenue left behind is vast.
  • As Good Life Resident on Social Security Tyler Bond told CNBC this week, “It’s really hard to just ignore the fact that scrapping the cap closes so much of the funding gap without needing to implement any other changes.”
  • Read Bond’s proposal to reform the payroll cap, Fixing the Tax Max: Restoring Social Security’s Wage Base, and check out the entire collection of Social Security reform ideas from our residents.
• “Trump accounts” don’t fight wealth inequality—they could make it worse. “The main mechanism for these accounts is active savings by parents, and that’s by nature regressive,” Roosevelt Senior Fellow Darrick Hamilton told The Guardian. “These accounts will actually increase wealth inequality because they favor people wealthy enough to have money to put into savings.”
• JD Vance’s childcare plan is to pit moms against each other. The vice president’s proposal to give stay-at-home parents a subsidy sounds like a good idea—until you learn that the funds would be stripped from childcare subsidies for low-income working parents, who are often unmarried mothers.
  • As Roosevelt Fellow Jessica Calarco writes for MSNOW, the Vance plan presents “marriage and at-home parenting as the right choices for women and the only ones deserving of government support.”

What we’re talking about