Why Are Palantir and OpenAI Scared of Alex Bores?

Summary of Why Are Palantir and OpenAI Scared of Alex Bores?

by New York Times Opinion

1h 32m•April 21, 2026
Listen to Episode

Overview of Why Are Palantir and OpenAI Scared of Alex Bores?

This New York Times Opinion interview profiles Alex Bores — a former Palantir employee, current New York State Assemblymember, and congressional candidate — focusing on his AI policy work (the RAISE Act), his politics and background, the controversy over his Palantir past, and why major AI/tech funders (OpenAI co‑founders, Andreessen Horowitz, Palantir co‑founders) are financing a super PAC attacking him. The conversation centers on the politics of AI regulation, concrete regulatory proposals Bores champions, the political economy of tech influence, and policy ideas to capture public benefit from AI rather than leave it solely to private firms.

Key topics discussed

  • Alex Bores’s origins: early union-influenced political formation, study of industrial & labor relations and computer science, campus activism.
  • Palantir tenure (2014–2019): work on federal civilian projects, internal resistance to ICE/ERO use of Palantir, reasons for leaving (ethical concerns vs. burnout allegations), and conflict with Palantir executives.
  • Political attack: a super PAC called Leading the Future (and related outfits) funded by major tech/VC players is spending millions to oppose Bores — reportedly to deter AI regulation by making an example of him.
  • RAISE Act (New York state): an early state AI bill Bores helped author and pass — baseline safety rules, public safety plans, reporting of critical safety incidents; stronger proposals (blocking release of unsafe models; mandated third‑party audits) were trimmed.
  • AI regulation philosophy: slow and govern, establish baseline standards, protect kids and workers, invest public capacity.
  • AI dividend and distribution: proposals to capture public value from AI (wealth taxes, token/usage taxes, out‑of‑the‑money warrants) to fund UBI/retraining and share benefits.
  • Data centers & grid policy: use data‑center investment as leverage to modernize the electric grid and accelerate renewable interconnection while requiring contributions to grid upgrades and high renewable shares.
  • Children, education, and pedagogy: age verification, mental‑health safeguards, limits on training on kids’ data, updating school assignments and assessment to reflect AI tools.
  • Government capacity and public AI agenda: build public compute (e.g., state GPU clusters), hire AI expertise into government, direct public research (drug discovery, alignment, safety).
  • Broader political-economy risk: tech money can intimidate legislators; bipartisan public concern about AI; advocacy for international diplomacy on AI arms‑race risks.

Main takeaways

  • Bores argues that powerful AI companies and investors are funding political campaigns to block or chill state and federal AI regulation — not merely to shape smart policy, but to deter legislators from acting at all.
  • The RAISE Act establishes an important baseline (safety plans, public reporting, defined critical safety incidents) but wound up weaker than proponents wanted because stronger accountability measures (release bans if models fail safety checks, mandatory third‑party audits) were removed.
  • Bores believes AI’s pace outstrips government capacity; urgent investment is needed in government expertise, public compute infrastructure, and regulatory frameworks to steer benefits toward public goods (drug discovery, public services) and away from unchecked displacement.
  • He advocates both harm‑reduction (safety testing, kids protections, reporting) and pro‑social strategies (AI dividend, public clusters, regulatory incentives) — and supports slowing development until alignment/safety gains are demonstrable.
  • The fight over Bores’s campaign is framed as a test case about whether Big Tech can buy de facto veto power over AI policy.

Notable specifics and policy proposals

  • RAISE Act (what it required and what was cut)
    • Required public safety plans that describe testing regimes and mitigation strategies.
    • Required government notification of defined critical safety incidents.
    • Initially included: (a) prohibition on releasing models that fail a company’s safety tests; (b) mandatory third‑party audits — both were removed during negotiations.
  • AI dividend (mechanisms to capture value for the public)
    • Wealth taxes or other progressive taxation.
    • Token/usage taxes on commercial AI usage, especially where labor is being replaced.
    • Government warrants in AI firms that pay off if companies achieve very large valuations (a way to participate in upside without expropriation after success).
    • Revenue would fund UBI, retraining, safety nets, and scale with AI adoption.
  • Data center policy
    • Require high renewable percentages for new data centers.
    • Charge data centers additional fees to fund grid upgrades and resilience; prioritize interconnection for centers meeting green and community standards.
  • Children and education
    • Age verification for certain AI interactions.
    • Protections for children’s training data and mental‑health evaluations for AI interactions.
    • Update pedagogy: rethink take‑home essays and assessment, require transparent creation logs/keystrokes, give teachers and parents visibility into student‑AI interactions.
  • Government capacity
    • Public GPU clusters (example: Empire AI, New York’s cluster) to enable universities and state agencies to experiment affordably.
    • Make it easier to hire and pay AI specialists in government; remove procedural barriers that leave agencies under‑resourced.
    • Use AI to modernize state systems (e.g., tax filing, benefit enrollment) and cut regulatory “cruft” to accelerate public benefit use cases.

Politics, power, and controversy

  • Leading the Future (super PAC) has major donors tied to OpenAI (some co‑founders), Andreessen Horowitz, and Palantir co‑founders — it is actively attacking Bores in his congressional race. Reported spend: $2.5M so far; they signaled they may spend much more.
  • The political objective (per Bores and reporting): to make an example of a legislator who succeeded in passing AI regulation, thereby deterring future regulation elsewhere.
  • Bores’s Palantir record is the campaign’s focal attack: he worked for Palantir 2014–2019 on federal civilian projects and left after disputes over Palantir’s refusal to add contractual guardrails preventing use for deportations. He denies the sexual‑harassment allegation reported in Bloomberg and says it was an internal pushback tactic.
  • Bores frames the conflict as emblematic of a larger choice: either let companies set the rules through money and influence, or have democratic institutions assert baseline rules for safety and public benefit.

Practical action items and recommendations (for policymakers and advocates)

  • Pass baseline AI safety regulation now: require public safety plans, testing, and incident reporting.
  • Restore/advance accountability measures: third‑party audits and mechanisms preventing release of models known to be unsafe.
  • Invest in government capacity: hire AI specialists, build public compute infrastructure, and modernize data systems to enable public use of AI (tax prefill, benefit enrollment automation, public health research).
  • Use data‑center approvals to enforce environmental and community benefits: high renewables thresholds, contributions to grid upgrades, and prioritized interconnection for green projects.
  • Develop policy to capture public value: consider token/usage taxes, warrants, or other mechanisms to fund an AI dividend and worker transition programs.
  • Protect children: age verification, data‑training limits, pedagogical updates, and parental/teacher transparency on student‑AI interactions.
  • Promote international diplomacy and verification frameworks to reduce an arms‑race dynamic in high‑capability AI.

Short bio — Alex Bores (as presented)

  • Education: industrial and labor relations at Cornell; computer science background.
  • Professional: worked at Palantir (2014–2019) focusing on federal civilian projects (DOJ, CDC, VA). Left amid concerns about Palantir’s contracts with ICE/ERO and internal refusals to add contractual guardrails.
  • Political: elected to New York State Assembly in 2022; recognized as an effective freshman legislator; co‑authored/passed the RAISE Act (one of the first state AI regulatory bills). Now running for Congress in New York’s 12th district.

Books recommended by Alex Bores

  • A Theory of Justice — John Rawls
  • World Eaters — Catherine Bracy
  • Bird by Bird — Anne Lamott

Bottom line

This episode sets up Alex Bores as a real‑world test of whether legislators can pass meaningful AI regulations in the face of concentrated tech/VC funding. Bores advocates a dual strategy: (1) immediate baseline safety rules and protections (especially for kids and the public), and (2) an affirmative public agenda to capture AI benefits via public compute, directed research, and mechanisms (AI dividend) to share gains and manage worker transitions. The central political threat he highlights is not just disagreement over policy details but the capacity of industry money to chill democratic oversight.