美國民主黨積極推動 AI 監管,黨內共識仍難形成
Democrats are fired up on AI. Consensus is harder.
美國民主黨表示將在新一屆國會推動 AI 監管,但黨內進步派與温和派、眾議院與參議院之間的分歧,令全面立法難以迅速推進。民主黨議員提出舉行 AI 公開聽證及傳召科技高層;參議院商務委員會民主黨首席成員 Maria Cantwell 正推動要求 AI 模型發佈前接受獨立測試的法案。
The News
The new Democratic majority in Congress knows AI is an issue ripe for the taking. A fight is already brewing over how aggressively to regulate it.
Tuesday’s election results saw more than 40 AI data center-related ballot measures passed that were designed to slow or stop growth outright. Conservative groups said it’s clear that voters are connecting economic woes with the unfettered growth of AI and an overly friendly environment for billionaire CEOs in President Donald Trump’s Washington.
However, divisions between the more progressive wing of the party and more business-friendly moderates, along with disagreements between the House and the Senate, will make it difficult for the newly empowered party to quickly pass any sweeping AI rules. And of course, Trump, who has thus far embraced a hands-off approach to AI regulation, has threatened to veto bills he doesn’t like.
“We know how important this is, and it is going to be a high priority for Democrats,” Rep. Greg Stanton, D-Ariz., a member of the New Democrat Coalition, told reporters on Wednesday. He acknowledged, however, that there is plenty of intra-party debate ahead.
The new Congress will usher in renewed pushes for bills to address AI safety and curb data center growth, as well as the formation of larger coalitions behind banning “superintelligence.”
But the lack of agreement around any one legislative policy push will make it difficult for Democrats to accomplish anything, even with larger majorities than they anticipated in both chambers.
“Not that we’ve reached a specific agreement on a specific policy,” Stanton said, adding that the caucus understands “how important it is” to address cybersecurity risks, protect kids online, and advance worker protections.
party projects optimism
Still, Democrats insist they’re in a position to deliver on AI guardrails regardless of divisions within their party.
One thing liberals and moderates can agree on: They’re already pledging to subpoena tech executives to grill them on safety risks and economic effects stemming from their technology.
“Many Democrats were negotiating with a Republican Party whose main AI policy goal last Congress was to eliminate all AI regulations in all states. Things have shifted drastically, such that even people like JD Vance are admitting they were completely wrong on the issues,” Rep. Greg Casar, D-Texas, told reporters Wednesday.
“The Democratic Party coming out of this election can prove to our voters that we are with them, and we are going to call the AI CEOs to testify under oath, not cash their checks,” he said, calling AI regulation a “top priority.”
Added Rep. Ted Lieu, D-Calif., in a separate conversation with reporters: “Two things will change next term. We will hold public hearings on AI. We will pass legislation and reasonable guardrails on AI.”
Rep. Josh Gottheimer, D-N.J., a member of the Problem Solvers Caucus, told Semafor that Democrats are in “surprisingly broad agreement” on AI.
“We agree that this technology is changing how Americans work and live. We also agree that we need commonsense guardrails, so that it works for families, workers, and kids — not against them,” Gottheimer said. “We’ll continue to have constructive discussions on how best to get there. I look forward to working with every member, new and returning, to get it done.”
New progressive voices on Capitol Hill, buoyed by midterm wins and stronger numbers, promise to make a lot of noise about delivering on AI.
“It’s going to take a lot of work to hammer out the details, but failure is not an option. AI poses a real and present danger to our economy and social fabric,” Will Lawrence, a Democrat who just won a House seat in Michigan, told Semafor. “There is a mandate for action from the American people.”
Room for Disagreement
Democrats see other ways to make an impact with their new power.
One Democratic Capitol Hill staffer told Semafor Democrats can now “subpoena the hell” out of Trump’s executive order on AI from June that has never fully been released publicly, along with using their subpoena power to question the administration’s use of export controls and the influence of tech players like Meta CEO Mark Zuckerberg on AI policy.
The staffer said he expects Senate leadership to cater to moderates and be reluctant to “go too far in one direction,” since AI is still a new political issue on which the party’s platform isn’t yet defined. Democrats are shifting left on AI, the staffer argued, but they doubted Democrats in both chambers would broadly support proposals like Casar’s bill to pause the development of advanced AI models.
Notable
- Abdul El-Sayed, the incoming senator from Michigan, has been among the most vocal Democrats raising alarm about AI on the campaign trail.
- Any major AI safety and testing legislation in the Senate would come from the Senate Commerce Committee. Sen. Maria Cantwell, D-Wash., the committee’s top Democrat, has been pushing for a bill that would mandate independent testing of AI models before their release.
來源:Semafor · semafor.com