Bernie's AI Tax Will Backfire. Hard.

Bernie Sanders' proposed AI sovereign wealth fund includes an alarming core tenet: a one-time, 50% tax paid in company stock from leading AI labs like OpenAI, Anthropic, and xAI. This isn't a tax on profits or cash; it's an outright expropriation of equity, effectively a "seizure" designed to grant a newly created federal fund a controlling stake in these crucial firms.

Proponents justify this unprecedented move with two primary arguments. First, generative AI models are built on humanity's collective knowledge, creative work, and data—often without compensation—and heavily leverage publicly funded research. Second, a Senate office report warns of profound societal disruption, projecting up to 97 million U.S. jobs displaced over the next decade, necessitating public benefit from this wealth creation.

Most controversially, the proposal mandates "democratic control." This means government would receive voting shares and equal board representation in the named labs, crucially gaining veto power over corporate decisions deemed harmful to citizens. The stated intent is to prevent a handful of unelected executives from unilaterally dictating societal transformation behind closed doors, ensuring public oversight of powerful AI development and its societal impact.

Sanders' plan faces an immediate constitutional quagmire. The Fifth Amendment's Takings Clause prohibits the government from taking private property for public use without just compensation. Seizing 50% of a company's equity—a direct expropriation of wealth—would undoubtedly trigger a protracted legal battle, effectively paralyzing the entire initiative in federal courts for years. This isn't a tax on profits; it's a direct seizure of ownership, presenting a fundamental challenge to property rights.

Economic realities also doom the proposal. Such an aggressive equity grab would trigger massive capital flight, driving founders, top engineering talent, and essential venture capital away from the United States. Highly mobile AI firms would simply reincorporate or relocate their operations to more favorable jurisdictions, like Ireland or the UAE, instantly undermining America's competitive edge in artificial intelligence.

Crucially, this plan fundamentally punishes risk and innovation. Why would investors fund high-risk, high-reward AI ventures in America if half their potential upside is immediately surrendered to the government? The immense disincentive would starve promising startups of necessary funding, stifling domestic AI development and ceding technological leadership to nations with more welcoming investment climates.

Proposal's targeting fundamentally misidentifies AI's financial engine. It focuses on high-cash-burn R&D labs like OpenAI and Anthropic, which are currently venture-backed and not yet consistently profitable. Imposing a one-time 50% equity 'seizure' on these nascent, capital-intensive entities risks stifling innovation and driving talent abroad, rather than capturing established wealth.

Meanwhile, the proposal ignores the industry's true, highly profitable giants. These include hyperscalers like Microsoft and Google, which sell the foundational cloud computing power and services AI models run on. Also overlooked are infrastructure 'picks and shovels' players: Nvidia, dominating the GPU market, and TSMC, fabricating advanced chips, both generating immense, consistent revenue from the AI boom.

Unprecedented conflict of interest would inevitably arise when the government assumes roles as a 50% owner, primary regulator, and major customer of the same AI companies. This entanglement grants the government undue power to block corporate decisions, distorts market competition, and creates significant risk of political capture. For further details on the proposal's intent and mechanics, readers can review the Bernie Sanders Op-Ed: The American AI Sovereign Wealth Fund Act.

Bernie Sanders' 50% AI tax, initially floated as an op-ed, functions less as a serious legislative proposal and more as a messaging bill. Its primary intent is to shift the Overton Window, sparking public discourse on how government can capture wealth from artificial intelligence and redefine the social contract. This approach acknowledges the public's role in training data and foundational research.

A more practical model for public wealth generation exists, proven by successful funds like Norway's Government Pension Fund Global and Alaska's Permanent Fund. These entities acquire passive, non-controlling stakes, typically capped around 10%, by purchasing shares on the open market. Such a strategy prevents government overreach and avoids market distortion, aligning public interest with long-term economic growth rather than immediate seizure.

Instead of a one-time stock seizure from a few high-burn R&D labs, a sustainable public AI fund requires a broader, gradual funding mechanism. Propose a small, ongoing levy across the entire AI ecosystem. This could target areas like compute resources, data usage, or windfall profits from established AI applications. Such a system ensures equitable contributions without crippling innovation or driving capital and talent to foreign jurisdictions.

What is Bernie Sanders' AI sovereign wealth fund proposal?

It's a proposal for a one-time 50% tax on major AI companies like OpenAI and Anthropic, paid in company stock. The plan also gives the government board seats and veto power over corporate decisions.

What are the main criticisms of the plan?

Critics argue it's likely unconstitutional (violating the Takings Clause), would cause capital flight of talent and investment, and punishes innovation by seizing equity from successful, high-risk ventures.

Is there a better way to create a public AI wealth fund?

Yes. A more viable model would involve a broad, gradual tax on the entire AI ecosystem's windfall profits, with the government taking passive, non-controlling stakes, similar to Norway's successful sovereign wealth fund.

Why does the plan target companies like OpenAI and not Google?

The proposal focuses on visible frontier AI labs, likely for political impact. However, this ignores the highly profitable hyperscalers and infrastructure companies that are the true financial winners of the AI boom.