I spent 15 years as a public defender in Manhattan watching what happens when powerful systems make life-altering decisions about people who have no ability to fight back. So when I tell you that AI systems are right now screening resumes, scoring video interviews, and deciding who gets a loan or an apartment—with almost no accountability when they discriminate—I want you to understand that I recognize the pattern.
The decision gets made. The person on the receiving end never learns why. And the entity responsible gets to shrug and point at a process.
Recently, I sat down with Congresswoman Sara Jacobs of California to talk about her newly introduced bill, the Sectoral AI Governance Act. I’ve watched a lot of AI legislation come and go, and most of it falls into one of two traps: it’s either so sweeping it has no chance of passing, or so watered down it protects no one. This bill threads the needle in a way I haven’t seen before, and I want to walk you through why.
The laws already exist. The enforcement doesn’t.
Here’s the core insight of the bill, and it’s deceptively simple.
We already have the Americans with Disabilities Act. We already have laws against employment discrimination. We already have fair housing laws and fair lending laws and civil rights protections that took generations of struggle to win. Those laws don’t stop applying just because a company routes its decisions through an algorithm.
But in practice, they might as well.
Rep. Jacobs gave me an example that stuck with me. Imagine an AI video interview system—the kind that’s already screening candidates at companies across the country—that systematically downgrades applicants based on their facial expressions. That system could be discriminating against people with disabilities in ways that have nothing whatsoever to do with their qualifications for the job. That’s a straight-up ADA violation.
Now try to enforce it. The EEOC, the agency responsible for policing employment discrimination, doesn’t currently have the framework or the tools to require an employer to verify that its AI hiring system isn’t producing discriminatory outcomes before deploying it. The law is on the books. The violation is happening. And the agency tasked with stopping it is essentially unarmed.
The Sectoral AI Governance Act fixes this by going sector by sector. Instead of creating some brand new AI super-regulator, it takes the agencies that already enforce our existing laws—the EEOC for employment, and the agencies covering housing, lending, and civil rights—and gives each of them the tools to enforce those laws when AI is making the decisions instead of humans.
As Rep. Jacobs put it: just because something is done by AI doesn’t mean anyone gets to disregard the laws we’ve all already agreed on.
The killer robot distraction
One of the things I appreciated most in our conversation was Jacobs pushing back on how the AI debate gets framed.
So much of the public conversation about AI is consumed by the big existential scenarios. Will AI take over the world. Will it kill us all. Jacobs was candid that those are fair concerns. But she worries—and I share this worry—that focusing only on the hypothetical apocalypse ignores the very real harms AI is doing in people’s lives right now.
The job you didn’t get. The apartment application that was denied. The loan that was rejected. Somewhere, right now, an algorithm is making one of those decisions about someone, and that person will never know why, and under current enforcement realities, no one will ever have to answer for it.
Those harms don’t make headlines the way killer robots do. They just quietly reshape people’s lives.
Jacobs was also clear that this bill won’t solve everything, and she’s not pretending it will. She has a Model Transparency Act in the works to create rules around the foundational models themselves, plus a proposed commission to study AI’s impacts on the economy. She sees this as a multi-level fight. But she called the Sectoral AI Governance Act the most ambitious and practical bill yet introduced to address the way AI shows up in our daily lives, and having read it, I think that’s right.
What you can do today
The bill was recently introduced, which means it currently has zero co-sponsors. That’s where you come in, and this is one of those rare moments where a phone call genuinely matters.
Call your member of Congress at (202) 224-3121 and ask them to co-sponsor the Sectoral AI Governance Act. The Capitol switchboard will connect you to your representative’s office. The call takes two minutes. You say you’re a constituent, you say you want them to co-sponsor the Sectoral AI Governance Act, you say thank you, you hang up. Congressional offices track these calls, and early co-sponsorship momentum is what determines whether a bill builds a coalition or dies quietly in committee.
And beyond this one bill: keep bringing up AI when you talk to your representatives. Jacobs told me something that should alarm all of us, which is that many of her colleagues simply aren’t thinking deeply about this and don’t understand how these systems work. Members of Congress pay attention to what their constituents raise. If nobody’s calling about AI, nobody in Congress feels any urgency to act on it.
The technology is moving fast. The harms are already here. The laws to address them already exist. All that’s missing is the enforcement—and that part is up to Congress, which means that part is up to us.
Watch the full interview above, and if this was useful to you, share it with someone who’s job or apartment hunting right now. They deserve to know what’s screening them.
—Eliza
