New York Assemblymember Alex Bores is launching a $30 million political effort aimed at making AI regulation a defining issue for Democratic candidates by 2028. The new organization, called Who Decides, plans to push a common artificial intelligence policy platform across presidential, statewide and down-ballot races.
The campaign matters because it signals that AI regulation is becoming a major political battleground, not just a technology debate. For investors, that raises the likelihood of more formal rules for frontier AI developers, with potential implications for compliance costs, product timelines and competitive positioning across the sector.
The initiative also underscores a broader shift: concerns about AI safety, labor disruption and democratic accountability are being translated into electoral strategy. If that effort gains traction in key states, it could influence both federal regulation and the valuation outlook for leading AI companies.
Key Facts
- Alex Bores and former chief of staff Anna Myers are launching Who Decides with a stated $30 million goal focused on AI policy ahead of the 2028 elections.
- The group plans to operate in 11 states: South Carolina, Nevada, New Hampshire, New Mexico, Michigan, Virginia, Arizona, Georgia, North Carolina, Pennsylvania and Wisconsin.
- A separate report summarized in the article said the nonprofit has raised $10 million so far and plans to seek another $20 million in 2027.
- Bores previously received support from a political action committee funded with a $20 million donation from Anthropic, alongside $401,250 in employee political contributions backing his campaign.
- The organization says it will reject corporate money and contributions from senior executives at frontier AI companies.
AI regulation campaign
The launch of Who Decides shows how quickly AI regulation campaign politics are evolving. Bores is trying to turn a fragmented mix of public anxieties into a coherent policy agenda, centered on issues such as published safety plans, critical incident reporting and independent testing of advanced AI models. Those ideas mirror several proposals already circulating in policy and industry circles, but the new effort is designed to move them into campaign platforms and, eventually, governing priorities.
The timing is notable. The AI industry has faced growing scrutiny over the speed of model development, the concentration of power among a handful of companies and the possibility that advanced systems could disrupt labor markets or create broader safety risks. At the same time, political leaders are split over how aggressive Washington should be. Some favor faster federal intervention, while others argue that overregulation could slow U.S. innovation and strengthen foreign rivals, especially China.
Bores’ argument is that AI governance should not be left primarily to industry insiders. The organization plans to use polling, conferences, questionnaires, forums and endorsements to shape candidate positions. That means AI policy may become more visible in primaries as well as the general election, particularly in swing states where economic dislocation, education, labor and consumer protection concerns can resonate beyond the technology sector.
“We don’t let five people write the laws for hundreds of millions of Americans. There is no reason to let them write the rules for AI.”
How the effort could shape policy
The group’s strategy points to a more organized push for federal standards that could be actionable early in a new administration. Its launch materials frame AI as an issue that a future Democratic president could address within the first 100 days in office. That suggests a near-term focus on executive actions, agency enforcement, procurement standards and disclosure frameworks, even if broader legislation takes longer to pass.
Bores also brings state-level experience to the effort. He helped pass New York’s RAISE Act, which requires major frontier AI developers to publish safety plans and report serious incidents. If similar concepts spread nationally, large AI labs could face a more structured compliance environment, while smaller firms may need to prepare for indirect effects through vendors, cloud partners and enterprise customers.
Implications for Investors
For investors, the biggest takeaway is that AI regulation is becoming a material policy variable rather than a distant headline risk. Companies building frontier models may need to budget for higher compliance spending, third-party testing, reporting systems and governance staff. Those added costs could favor the largest players, which have the balance sheets to absorb regulation, but they could also constrain speed-to-market and reduce flexibility in product deployment.
There is also a political-risk dimension. If AI becomes a defining campaign issue in 2028, investor sentiment toward leading private and public AI-linked companies could increasingly hinge on election probabilities, party platforms and state-level activism. Market participants should watch whether proposed rules stay focused on transparency and safety or expand into licensing, liability, labor protections and antitrust oversight. The broader and more prescriptive the agenda becomes, the greater the impact on margins and long-term capital allocation.
At the same time, regulation could create opportunity. Firms offering AI auditing, model evaluation, cybersecurity, compliance software and governance tools may benefit from rising demand if disclosure and testing requirements become standard. Cloud providers, consultants and enterprise software vendors that help customers manage AI risk may also gain. In that sense, political momentum around AI safety could reshape not only the economics of model creators but also the wider ecosystem that supports responsible deployment.
The next phase will depend on fundraising, candidate uptake and whether AI remains a top-tier voter concern through 2027 and 2028. For now, Bores’ initiative is another sign that the commercial future of artificial intelligence will be shaped as much by politics and regulation as by computing power and model performance.