Policy Commentary: Bernie Sanders’ Proposed AI Sovereign Wealth Fund Act
Date: August 11, 2026
Read the full PDF here: Policy Commentary: Bernie Sanders’ Proposed AI Sovereign Wealth Fund Act
Introduction
On June 18, 2026, Bernie Sanders (I-Vt.) introduced a bill to the US Senate proposing the creation of a sovereign wealth fund financed by the artificial intelligence (AI) industry (Bill S.4825). This bill has strong rationale and reasoning behind its creation, as well as a clear vision and expectation for future results.
In short, Bill S.4825 calls for:
(i) A one-time 50% tax on the stock of the largest AI companies. This monetary sum would be deposited into a sovereign wealth fund.
(ii) A 5% annual dividend from this fund that provides a direct payment to every American, with the first payment estimated to start at $1,000.
(iii) The creation of an Independent Commission for Democratic AI to manage the sovereign wealth fund in the public interest.
These three mechanisms work in tandem to give the public a direct ownership stake in the largest AI companies in America. While Bill S.4825 could be further bolstered by highlighting breaches of data privacy and strengthened by including a tailored definition of artificial intelligence, the bill proposes a sound and comprehensive system for democratic oversight of the AI industry and distribution of AI-created wealth amongst all Americans.
Intentionality
Sanders’s bill, S.4825, has strong reasoning, intentionality, and precedence supporting it. The bill begins by contextualizing the purpose and utility of sovereign wealth funds, outlining examples of existing successful funds around the world. The bill explains that six states in the US handle over $300 billion in sovereign wealth fund assets; these assets result in annual checks ranging from $1,000 to $3,000 sent to eligible citizens, as well as $5 billion sums being distributed to public schools.
S.4825 succinctly describes how and why AI is inherently a publicly supported resource. The bill explains that AI models are trained using the “collective intelligence” of humanity. This collective intelligence is made up of the knowledge and outputs of groups and individuals, including but not limited to authors, artists, research teams, and companies.
There is opportunity here to strengthen Sanders’ argument regarding AI’s reliance on and origins in the contributions of humanity – societies, communities, and individuals – by highlighting the many instances in which surveillance technologies have been instrumentalized to populate AI data training sets. Recent history abounds with examples of technology companies collecting information on citizens’ behaviors, biometrics, social media profiles, web-browsing metadata, Wi-Fi routers, IP addresses, GPS signals, purchasing patterns, health data, and more, often without an individual’s express consent. Many of these instances of nonconsensual data collection have been connected to AI training and development programs. The bill can strengthen its point here even further, driving home the notion that the AI industry is not only built from the collective intentional work of humanity, but also built from data that has been misappropriated on numerous occasions.
Scope
The scope of the AI Sovereign Wealth Fund bill is well defined and the inclusion of industries adjacent to AI algorithms themselves reflects comprehensive forethought. The bill addresses AI data centers, AI computing infrastructure, AI services, and advanced robotics, and includes these practice areas under the umbrella of applicable AI trades or businesses that the bill aims to address. This scope and range encompass multiple parts of the AI lifecycle without integrating all aspects of computing in general.
As comprehensive as the umbrella of “AI trade or business” is, the bill lacks a definition of AI in general; this exposes a vulnerability in the bill that may hinder enforcement of the bill’s goals. It is true that there are few agreed-upon definitions of artificial intelligence; the technology is dynamically changing and rapidly advancing. If anything, the fluidity of the concept of “artificial intelligence” makes it crucial that this bill explicitly outlines a definition of what technologies are included under this phrase. A detailed definition of AI should be put forth in this bill so that companies cannot use semantics to change the name of the services they offer in order to no longer fall under the “AI trade or business” umbrella as specified. This is a protective enforcement mechanism that mirrors the bill’s stipulations that discourage companies in the AI industry from moving offshore to evade the 50% tax outlined in the bill.
Bill S.4825 does not have to create a new definition of AI, but can adopt an existing definition. The Internal Revenue Service (IRS) Office of Professional Responsibility (OPR) has issued a statement classifying AI as “the use of machines in a way that mimics human cognitive skills, including judgment, perception, and prioritization” [see issue number 2026-19]. This definition can be expanded upon to capture a more comprehensive representation of the technology, as exemplified below in the definition outlined by the European Union High Level Expert Group:
“The term AI refers to:
(i) Systems that display intelligent behavior by analyzing their environment and taking actions – with some degree of autonomy – to achieve specific goals.
(ii) Systems that utilize machine learning to continuously or periodically change their patterns of behavior, update their systems of reasoning, or adjust their outcomes in order to adapt to novel scenarios.
(iii) Systems that are built with the intention of enabling computers and machines to simulate human learning, comprehension, problem solving, decision making, creativity and autonomy.”
Methodology
One of the greatest strengths of Bill S.4825 is the clear mechanism of AI-related financial empowerment that it offers Americans. This is exemplified by the proposed 50% equity tax on companies in the AI industry, and the subsequent rollout and handling of funds by an independent commission. Economic equality is a foundation of democracy, and preventing the concentration of wealth is of utmost importance in preserving our democratic identity.
However, it must be clarified how the wealth fund and its board of commissioners will shape the AI industry through corporate governance. The bill states that the commissioners will exercise all voting and governance rights attached to the equity interests held in the American A.I. Sovereign Wealth Fund, indicating that there will be the possibility of the commissioners using voting shares in companies to block or promote corporate decisions, depending on their effect on the American public. This creates a potential conflict of interest, where the commissioners have to manage the fund’s balance while simultaneously making decisions on the future of AI in the boardroom. Further clarification is needed here to ensure that fund commissioners will not use their voting interests to promote decisions that may increase a company’s value, thereby increasing the value of the fund, even if it comes at the expense of the public’s benefit. The bill may be strengthened by a further separation the fund’s financial and managerial duties.
The bill begins to address this ambiguity in the role of the commissioners. The bill features language explaining how the fund will be used to give payouts to Americans and the fact that the fund aims to pursue “decent and dignified standard of living, including health care, education, housing, and a healthy and habitable environment, in such manner as Congress may provide.” The bill specifies that the commissioners have a duty to “manage the American A.I. Sovereign Wealth Fund, in consultation with the Secretary of Labor, in such manner as to effectively promote the goals of worker welfare, public safety, fair competition among applicable AI companies, environmental sustainability, and financial solvency. Lastly, SEC. 9512 of the bill requires commissioners to publicly disclose how they exercised the voting rights in each applicable AI company. These clarifications begin to delineate roles and boundaries for the commissioners in their dual duties to both the American public and the sovereign wealth fund.
It must nevertheless be noted that sovereign wealth funds hold an immense amount of capital that can shape industry behavior and external markets. Bill S.4825 may benefit from further specification around how the fund will invest its assets and the stipulations for ensuring that the fund’s assets will continuously be used to promote the best interests of the American people, not just the best interests of the fund’s balance.
Conclusion
Senator Sanders’ proposed AI Sovereign Wealth Fund bill outlines a promising pathway towards increasing public ownership in AI. Artificial intelligence systems continue to be embedded throughout society and continue to be built from human activity. Even as the effects of AI become more profound in the livelihoods of Americans, the public remains increasingly separated from the decision-making regarding the future of technology. Additionally, the gap between the wealth of the public and the wealth of AI developers increases with each passing day. It is crucial that the public has adequate representation in the AI industry and has the means to influence how the technologies of the future will be developed; Bill S.4825 begins to address financial parity and public representation in AI development.
Notes
See Csupo v. Alphabet Inc., No. 19CV352557 (Cal. Super. Ct. Santa Clara Cnty., jury verdict July 1, 2025); Clearview AI, Inc., Consumer Privacy Litigation, No. 1:21-cv-00135, MDL No. 2967 (N.D. Ill., final approval Mar. 20, 2025); Facebook Biometric Information Privacy Litigation, No. 3:15-cv-03747 (N.D. Cal., final approval Feb. 26, 2021); United States v. Amazon.com, Inc., No. 2:23-cv-00811 (W.D. Wash., filed May 31, 2023); Dinerstein v. Google, LLC, 73 F.4th 502 (7th Cir. 2023); Vance v. International Business Machines Corp., No. 1:20-cv-00577 (N.D. Ill.).
See IRS OPR issue number 2026-19.