Andrew Yang's $300 Billion Data Dividend Pitch Fuels Bitcoin (BTC) Payout Debate

Andrew Yang's $300 Billion Data Claim

Andrew Yang put data ownership back at the center of the AI debate on September 27, telling AI optimists on the Jubilee Media program “Surrounded” that companies profiting from models trained on personal data and creative works owe direct payment to the people behind that material. The core of his case is a “Data Dividend”: any firm that earns revenue from AI systems built on user data or user content should return a share of that revenue to the data providers and creators who supplied it. Yang pegged the commercial value of personal data at more than $300 billion a year, arguing that almost none of that value flows back to the public. He cited his own example as proof of concept — his publisher received a $2,500 licensing payment after an AI company absorbed the contents of one of his books into its training corpus — and argued that every content creator deserves equivalent treatment when a model consumes their work. His labor-market warning was sharper still. Yang argued AI may eliminate existing jobs faster than it creates new ones, hitting middle-skill white-collar roles first: lawyers, accountants and data analysts. As companies redirect capital away from offices and hiring toward servers and data centers, he said, capital, data and computing power pool inside a small group of technology firms, hardening a winner-take-all structure. He noted that the businesses monetizing this data span the entire economy — consumer-data-driven retail giants like retail giant Walmart and workflow platforms such as automation vendor ServiceNow — yet none of them currently pay the individuals whose data underpins their AI products.

A Universal Basic Income Funded by a VAT

Yang's remedy package pairs the Data Dividend with universal basic income. Under his framing, dividends and new taxes on technology companies would bankroll basic income payments, and recipients spending those funds on local consumption would sustain regional economic circulation. On the revenue side, he called for introducing a value-added tax, arguing that conventional corporate income tax alone cannot stop multinational technology firms from shifting profits across borders through accounting structures. A VAT would instead capture value at every commercial step — digital services, advertising clicks, data transactions — leaving less room for avoidance. He also pointed to sovereign-wealth-fund-style structures, or a weighted index of technology companies, as models for pooling AI-driven returns and distributing a universal dividend to the public. The supply chain behind AI explains why the money at stake is so large: chipmakers like Samsung Electronics sell the compute that trains the models, while software and content platforms like Adobe supply the licensed creative material increasingly fed into training corpora — layers of value Yang wants taxed and shared. He was candid about the open questions: how to value individual data, verify who actually holds the rights, structure corporate payouts and define how much of a creator's work may be licensed for AI training in the first place. By the end of the exchange, the debate had narrowed to a single question — not whether AI development should slow down, but who captures the gains it produces. Readers tracking the market in real time can follow live spot and futures prices on Gate.

Why Bitcoin (BTC) Enters the Picture

Our reading at COINOTAG: the debate has shifted from whether AI should be paused to who keeps the proceeds — an ownership question crypto has always answered architecturally. Paying millions of creators for granular, per-use licensing implies programmable, neutral settlement at internet scale, a role decentralized settlement layers were built for, with Bitcoin (BTC) the asset most often raised for that role — though Yang himself named no cryptocurrency. The broadcast, which aired September 27, is the primary record here: the Data Dividend remains a policy pitch, not legislation, and no implementation mechanism yet exists.