Andrew Yang just dropped his AI tax pitch on CNBC’s Power Lunch. Again. The 2020 presidential candidate wants the government to tax artificial intelligence instead of payroll. He’s not alone. Anthropic’s CEO Dario Amodei floated a 3% AI revenue tax in 2025. Bridgewater’s executives are now pushing a tokenized version. The narrative is building: tax the machine, not the worker.
Yang’s logic is simple. Firms skip payroll taxes and healthcare costs by choosing AI over new hires. Tax the AI, and the cost equation flips. He’s been here before—his 2020 campaign was built on automation warnings and a Universal Basic Income called the Freedom Dividend. He also backed crypto adoption and clearer digital asset rules. Now he’s championing an AI tax as a replacement for payroll tax. The revenue? Direct checks to displaced workers. Retraining programs? He calls them failures, citing coal miners and warehouse staff.
But the data backs the fear. A CNBC and Generation Lab survey of Americans aged 18-34 found 45% expect AI to hurt their careers. Only 10% see it as a help. Bridgewater’s Greg Jensen and Nir Bar Dea estimated 18% of current US jobs could be displaced within five years. Customer service alone employs 2.9 million Americans. The writing is on the wall. But the solution? A tax.
Beacon chain stable. Fragility remains.
Let’s audit the proposal. Amodei’s 3% revenue tax applies each time a model generates revenue. Yang wants it broad. The idea is to force firms to weigh AI costs against payroll costs. On paper, it sounds like a level playing field. In practice, it’s a mess. AI revenue is not like payroll. Payroll is a fixed cost per employee. AI revenue is variable, scalable, and often opaque. How do you tax a model that generates revenue through unpredictable inference calls? The compliance burden alone would crush small startups. The big players—OpenAI, Anthropic, Google—can afford the accountants. The rest will find workarounds.
That’s where crypto enters the frame. Tokenized AI services already exist. Projects like Bittensor, Render Network, and Akash Network allow users to pay for compute or model inference with tokens. Revenue is not fiat—it’s crypto. Taxing that revenue requires tracking on-chain transactions, which is possible but messy. The IRS has no clear guidance on taxing token revenue from AI models. The proposal ignores this entirely.
Based on my audit experience with AI-crypto hybrid projects, the tax logic is flawed from the start. These projects structure revenue as token emissions, not direct sales. A model generates revenue, but the revenue is often in the form of newly minted tokens that are unstable and illiquid. Taxing that at 3% is like taxing a derivative on a derivative. The real revenue is in the token appreciation, not the transaction. The tax base doesn’t exist where they think it does.
Audit passed. Trust failed.
Now the contrarian angle. The AI tax push, for all its flaws, might actually accelerate crypto adoption. Here’s why. If the government taxes AI revenue, firms will look for ways to avoid the tax. One way is to shift to decentralized AI platforms where revenue is generated and distributed via smart contracts. No central entity to tax. Another is to use token-based compensation for workers—paying humans in crypto, not payroll. This shifts the tax burden from payroll to capital gains, which is often lower. Yang’s proposal could inadvertently create a tax arbitrage that favors crypto-native AI businesses.
Bridgewater’s token tax proposal is a hint. They suggest an AI token tax, not a revenue tax. That’s a different animal. A token tax would hit the issuance and transfer of AI-related tokens. That would directly impact the crypto AI sector. But it’s also easier to implement—just tax every token transaction. The problem is that it would kill innovation. Small AI token projects would be taxed out of existence. The big ones would pivot to off-chain structures.
The unspoken truth is that the entire AI tax debate is a distraction. The real issue is that AI is creating a new class of digital labor that doesn’t fit existing tax frameworks. Crypto can solve this—smart contract-based compensation, automated tax withholding, transparent revenue sharing. But the government’s approach is backward. They’re trying to fit a square peg into a round hole. The code is already there. The question is whether regulators will let it scale.
NFT floor? More like NFT fiction.
I’ve seen this pattern before. In 2021, when NFT royalties were the hot topic, OpenSea’s surrender killed the creator economy. The same thing is happening now. The AI tax debate is a policy fiction that ignores the technical reality. The real impact will be on the ground—startups moving to decentralized platforms, token-based compensation, and a new wave of regulatory arbitrage. The winners will be the projects that can navigate this chaos.
Forward-looking thought: Watch for AI token projects that explicitly market themselves as ‘tax-optimized’ or ‘labor-displacement-proof’. The narrative is shifting from UBI to token-based incentives. The code is already there. The question is whether regulators will let it scale. Yang’s tax pitch is a warning shot, not a solution. The market will price the risk. I’m watching the on-chain data.
Fast news requires faster fact-checking. This one’s still unfolding.
