Elizabeth Warren

Latest official personal statement. Calls frontier models dangerous without adequate safeguards and demands an immediate pause in advanced development while lawmakers and regulators establish protection. Rejects voluntary pacing and weakening antitrust as a response. Names cyberattack, economic crisis and national-security disaster as reasons for urgent legislative action.

Full authored TIME op-ed reproduced by her office. Recognizes tremendous promise but calls for sharing gains, changing tax incentives that favor replacing workers, wealth and corporate taxation, and an energy-use excise tax. Conditional mass-displacement responses include health care, training, a jobs guarantee and stronger unemployment insurance. Does not claim those scenarios are inevitable.

Full prepared remarks inspected. Says AI has enormous potential and expressly declines to forecast its future inventions or capabilities. Sees an investment bubble and a possible debt-driven financial crisis that could affect savings, credit and unrelated workers. Proposes structural financial safeguards, a digital regulator, taxation and executive accountability; the crash remains conditional.

Bipartisan demand for mandatory annual disclosure of energy consumption, electricity pricing, deposits, grid-upgrade costs and their allocation. Argues voluntary industry promises cannot be enforced without comprehensive data. Accountability and protecting household ratepayers, not a forecast of unlimited AI energy growth, are the personal policy commitments.

Her letters question apparent retaliation against Anthropic and whether OpenAI’s Pentagon contract permits domestic surveillance or autonomous targeting without adequate protection. Demands meaningful constraints to prevent civilian harm. Allegations and doubts about contract safeguards are recorded as her concerns, not settled findings.

Requests financial transparency and details on possible taxpayer support. Objects to privatizing profits while socializing failures; acknowledges tax credits and loan guarantees can legitimately support critical industries and jobs. Financing concerns do not establish that AI is useless, and do not by themselves prove a bailout will occur.

Bipartisan bill targets vendor lock-in, competitive contracting, government data rights, barriers for smaller suppliers and consideration of multiple cloud providers. Supports useful public-sector innovation while resisting monopoly dependence. Sponsoring the bill is a policy preference, not evidence its provisions became law.

Questions whether Google-Anthropic and Microsoft-OpenAI partnerships evade antitrust scrutiny, consolidate talent and computing, and increase switching costs. Seeks information about reduced choice, prices and innovation. The letter’s concerns are not an adjudicated finding that either partnership is illegal.