Alex Bores wants to turn AI into a party-wide issue
New York Assemblymember Alex Bores is rolling out a new effort called Who Decides, and the pitch is pretty blunt: Democrats should stop treating artificial intelligence as a side conversation for policy nerds and start treating it like a message they can carry into the next presidential cycle. Bores has already made a name for himself by pushing back on unchecked AI, so this move feels less like a hobby project and more like the next step in a running political argument he’s been having in public.
The wager here’s simple enough to fit on a bumper sticker, which is probably part of the appeal. If AI is going to keep shaping work, elections, education, entertainment and a whole lot of daily digital life, then Democrats can either react to it in fragments or decide how they want to talk about it together. Who Decides is being framed as the second option: not a one-off campaign message, not a paper statement that gets trotted out once and forgotten, but a shared answer that candidates could actually use.
The gamble is that Democrats can move from scattered anxiety about AI to a real message before the issue hardens into somebody else’s brand.
That timing matters because AI policy has left the conference-room phase. It’s no longer just a niche fight between lawmakers, engineers and the occasional very stressed academic. It’s now showing up in power and politics, in campaign rhetoric, in labor fights, and in the broader churn of digital culture. Voters may not know the difference between model training, inference and whatever product demo is getting dragged across the internet this week, but they do know when a technology starts rearranging jobs, media and everyday routines. The political class has noticed, too, which is why AI is now getting folded into tech news, election strategy, and the usual scramble to define what the future’s supposed to look like.
Bores seems to be betting that frustration about AI can be turned into something sturdier than a news cycle reaction. That’s the interesting part. Plenty of politicians can condemn a technology when it’s causing trouble. Fewer can turn that annoyance into a coalition with a clear lane through party politics. His effort suggests he wants Democrats to get past the usual split between people who want faster innovation, people who want stricter guardrails and people who mostly hope the whole subject will stop showing up in their inboxes. Good luck with that, by the way.
The bigger idea’s that AI may be moving into the same category as other issues that parties eventually had to decide what they believed, whether they liked the fight or not. Bores is trying to get there early, before the 2028 race locks in its own vocabulary. If he pulls it off, Democrats won’t just have a louder complaint about AI. They’ll have a line that sounds like a party, not a passing mood.
What Who Decides actually is — and what it is not
If the first part of this story is political ambition, this part’s plumbing. Who Decides is being sold as a serious fundraising operation, with a target in the tens of millions. That’s not the kind of number people toss around when they’re hoping for a few sympathetic checks and a nice launch post. It suggests lawyers, compliance people, staff and enough cash to make the project feel like a national outfit rather than a hobby with a mission statement.
What it doesn’t yet have is a clean public label. The group has not said whether it’ll function as a super PAC, a nonprofit advocacy shop, or some other legal arrangement. That distinction isn’t window dressing. A super PAC can raise and spend a lot of money, but it comes with rules about coordination and disclosures. And a nonprofit can do issue advocacy and public education under a different set of limits, but that route brings its own restrictions and headaches. Until the group chooses, donors and operatives are left reading the tea leaves and squinting a little.
In politics, the legal wrapper isn’t decoration. It decides what money can come in, what can go out, and how loudly the whole operation can speak.
There is, at least, a donate button on the launch site. The payment flow runs through a third-party fundraising platform commonly used by nonprofits, which gives the project a more finished feel even while the formal structure remains murky. In practice, that means someone has already done enough setup work to accept money, though not enough to explain exactly what kind of organization that money’s going into. For a brand-new political effort, that’s a pretty revealing middle ground.
The donor rules are just as telling. Who Decides says it won’t take money from major AI companies or from senior executives at frontier AI firms. That’s a fairly direct line in the sand. In plain terms, the group wants to argue for tougher AI policy without letting the companies building the tools write the biggest checks. It’s a sensible pitch if you’re trying to make an anti-capture argument before the first dollar clears. It also raises the obvious question of whether that promise can survive once the fundraising tempo picks up and the pressure to hit a giant target kicks in.
Bores has already made his views on AI plain elsewhere. In a public AI framework, he laid out a set of rules and guardrails that put him firmly in the regulation camp. He later pushed the same basic line in a letter on New York preemption, where the fight over whether states can set their own AI rules got a more pointed treatment. He has also taken that case onto the Assembly floor, where his remarks are preserved in a video from the New York State Assembly. The point here is not that he’s discovered AI policy yesterday. He’s been living in this lane for a while, and the new group looks built to turn that record into a donor pitch.
Still, there’s one awkward question hanging around the launch. Bores didn’t immediately address whether he helped raise money while serving in elected office. That matters because the line between public service and political fundraising can get fuzzy fast, especially when a sitting lawmaker’s starting a national advocacy project. Even when everything is legal, the optics can be a bit of a chew toy for opponents.
Anna Myers, his former chief of staff, is launching the project with him. That detail matters more than it might seem. A former chief of staff brings the boring but useful stuff that a political startup needs: calendars, contacts, message discipline, and some sense of which ideas survive contact with actual donors. If Bores is the face of the campaign, Myers looks like part of the operating setup. That’s usually how these things work when they’re trying to become more than a one-man crusade.
The whole setup has the feel of an operation that wants to look disciplined from day one. And the money target’s big. The donor ban is specific. The legal structure is still open. Even the fundraising portal has already been wired in. That combination says the group is trying to move quickly without pretending the paperwork doesn’t matter.
And paperwork does matter here. If Who Decides lands as a super PAC, it’ll have one kind of reach and one set of constraints. If it comes out as a nonprofit advocacy outfit, the playbook changes again. For now, it sits in that awkward pre-launch zone where the message’s public, the money chase is real and the final shape of the thing is still being sorted out in the background.
Why Bores thinks the political opening is real
The timing is doing a lot of work here. Alex Bores is launching a new AI politics machine just as the conversation around AI safety has gotten louder, sharper, and a little less theoretical. A year or two ago, you could still file AI under the “interesting but maybe tomorrow” drawer. That drawer is gone now. The alarm is coming from people who helped build the field, people who study it for a living, and people who have spent enough time around the machine to know where the loose screws are.
That matters for Bores because his pitch is not built around vague unease. He has spent his public life arguing that AI needs rules before the market, the hype cycle, or the next product race decides otherwise. On his AI framework PDF, his campaign platform, and even his Assembly biography, the same idea keeps coming back: AI should be governed, not just admired from a distance while everyone hopes for the best.
Bores is betting that AI anxiety has moved from a niche worry to something voters can actually use when they pick a side.
That bet is more reasonable now than it would’ve been a few years ago. The warnings have piled up. Former insiders have gone public with blunt arguments for stronger guardrails, and a steady stream of researchers and tech figures have said, in one form or another, that the industry’s moving faster than the rules can keep up. The tone’s changed too. It’s no longer just policy people in rooms with bad coffee. The issue keeps spilling into broader public debate, which is how an issue starts to look like politics instead of a seminar topic.
Both parties can feel that shift, even if they’re describing it in very different accents. Democrats are still sorting out whether AI should be framed as a worker issue, a consumer-protection issue, a national-security issue, or all three at once. Republicans, for their part, are trying to sound pro-innovation without sounding allergic to any rules whatsoever. Donald Trump’s pushed back against speed-run regulation, which gives his party’s officeholders a familiar cue: don’t talk like the federal government should tackle everything with a hammer and a conference call.
Bores built his House campaign around the opposite instinct. He made AI oversight one of the centerpieces of his run, along with support for the New York state RAISE Act, a bill designed to put safeguards around powerful systems before they’re used at scale with too little scrutiny. That was never the blandest message on the ballot, which may be part of why it attracted so much attention. It also made him a magnet for the tech policy crowd, the anti-regulation crowd, and the people who treat every AI vote like a proxy war over the future of the internet.
The money followed. His congressional race drew heavy spending from rival AI-aligned super PACs on both sides, turning a state-level contest into a surprisingly expensive argument about what the Democratic Party should say when the topic turns to machine intelligence. That kind of political fundraising doesn’t happen around a sleepy issue. It happens when people think the rulebook is still being written. Bores lost that primary to Micah Lasher, who later became the clear favorite for the seat, but the race seems to have done something useful for him anyway: it gave him a proof point. He can say he didn’t just post hot takes about AI regulation, he ran a real campaign on it and took the hits that came with the territory.
That’s the part that probably makes this new push more than a vanity project. Bores isn’t entering the AI debate as a neutral messenger. He’s entering it as someone who already took a swing, got outspent and decided the fight was still worth another round. For Democrats trying to figure out whether AI belongs in the same bucket as labor rights, consumer protection, and power politics, that gives him a story to tell that’s a little more durable than a press release. It’s also why his allies appear to think there’s room to turn scattered concern into something closer to a party line, before the next campaign season freezes everyone into slogans.
For now, the opening looks real enough. The warnings are louder, the tech crowd is no longer speaking with one polite voice, and the White House isn’t exactly inviting a rush to regulate. Bores sees a moment when AI policy can be sold as practical politics rather than abstraction. Whether voters will buy that pitch’s another question, but he’s not wrong to think the aisle has cleared a bit.
The 2028 play: states, polling, and a party consensus
If the launch sounds a bit ambitious for a new political group, that’s because it is. Who Decides is not trying to win a single news cycle or spit out one more talking point for cable TV to chew on. It’s aiming at the 2028 election, and it’s doing it the old-fashioned way: maps, memos, polling and a lot of retail politics.
The group says it’ll work in eleven states that matter both for the general election and for Democratic primaries. Pennsylvania, Michigan, Georgia and Wisconsin are part of the mix, which already tells you the theory of the case. These are places where Democrats need a message that can hold together suburban voters, union households, tech-curious professionals and the kind of primary voters who like to ask pointed questions at town halls. Not exactly a small audience, and not exactly a quiet one either.
The bet is that AI won’t stay a niche policy issue long enough for Democrats to ignore it, so the party might as well argue about it now, before the argument becomes a scramble.
That plan starts with polling. Who Decides wants to run surveys regularly at the national level and in the target states to see how voters react to different ways of talking about AI. That matters because “AI policy” can mean a hundred things depending on who’s speaking. To one voter. It sounds like safety rules for powerful systems. To another, it sounds like a job threat with a nicer haircut. To a third, it just sounds like politicians trying to sound smart about a thing they barely use beyond email drafts and bad photo filters.
Polling can sort some of that out. It can show whether voters respond better to warnings about safety, fairness, deepfake abuse, job disruption or concentrated corporate power. What stands out: it can also expose where the message falls flat. Politicians hate that part, naturally, but they need it. A broad concern about AI already exists. It just doesn’t arrive in one neat package. People worry about different pieces of the same machine, and those worries don’t always add up to a shared political line.
That’s where the rest of the operation comes in. The group plans conferences and local partnerships in the selected states, which suggests it wants more than a set of polling crosstabs sitting in a donor deck. Local organizations can do the unglamorous work that national campaigns usually fake with a press release. They know which elected officials are taken seriously, which union halls still draw a crowd, which campuses will show up for a panel, and which community groups can turn a dry policy discussion into an actual room with people in it.
That kind of state-by-state work also gives Democrats something they often struggle to build on new tech issues: repetition. If every candidate is making up their own language on AI, the party ends up with twelve partial answers and one confused voter. Copied and refined across battleground states, the party has a shot at sounding like it knows what it’s talking about, if the language gets tested. Ambitious? Sure. Slightly optimistic? Also yes. But it’s a cleaner strategy than hoping everyone magically lands on the same talking points after the first debate.
The stated aim is broad. Presidential contenders, Senate hopefuls, House candidates, state legislators, maybe even the folks running in races most people only hear about when the yard signs appear too early. Quite possibly, by 2028, Who Decides wants Democratic candidates at every level running on a shared AI safety agenda. That doesn’t mean every race will sound identical. It probably shouldn’t. A candidate in Wisconsin may talk about manufacturing and workplace disruption. Big difference. A candidate in Georgia may lean harder on fraud, scams and election manipulation. And a candidate in Pennsylvania may spend more time on labor and local industry. Same issue, different entry points.
And that may be the real test here. The concern around AI is already out there. The harder part’s turning it into a governing consensus before the politics harden into something messier. If Who Decides succeeds, Democrats may stop treating AI as a side topic that only shows up when a chatbot goes rogue or a deepfake hits the feed. They may start treating it as a party standard, the sort of issue that tells voters what the coalition believes about power, labor, safety and who gets to decide.
That’s a pretty big ask for a brand-new group. It’s also why this launch feels less like a one-off protest and more like the opening move in a fight over what kind of party Democrats want to be when AI is no longer the future, just the argument in the room.
