CFTC Chair Says Blockchain is the Only Cure for AI's Hallucination Problem

Michael Selig, the head honcho at the US Commodity Futures Trading Commission, has a novel prescription for the plague of AI-generated slop: a heavy dose of blockchain. His diagnosis? Use on-chain timestamps and identifiers to separate the authentic digital artifacts from the synthetic sludge, like a cryptographic purity test for your meme portfolio.

During a recent appearance on The Pomp Podcast, Selig tackled the market mayhem sparked by AI fakes. His proposed treatment was a classic crypto remedy: let the free market cook, with blockchain as the head chef. "If you can timestamp things and make sure there’s an identifier for each meme or AI generated posts, you can verify if it’s real or generated by AI," he stated, emphasizing that keeping this tech on US soil is non-negotiable for national sanity.

The regulator even went full degen with his prognosis, declaring, "you can’t have AI without blockchain." It's the kind of soundbite that would get a standing ovation at a crypto conference and a confused head-tilt in a Senate hearing.

When the topic shifted to policing autonomous AI trading bots, Selig prescribed the regulatory equivalent of microdosing. He fears the heavy hand of government could snuff out innovation before it even gets a chance to rug pull anyone. The CFTC's job, he argued, is to police the degens making the trades, not the anons building the tools they use.

This crusade for blockchain-based truth isn't some isolated thought experiment. It's part of a growing movement to use immutable ledgers to solve the authenticity crisis that AI just pumped full of steroids.

Proof-of-personhood protocols are a leading contender in this space. The poster child is Sam Altman's Worldcoin, which uses its World ID and some ethically-dubious eyeball-scanning to verify you're a human—a solution that trades privacy concerns for the ability to prove you're not a bot, a deal many find as comfortable as a 100x leverage position.

Just last month, Worldcoin rolled out AgentKit, a dev toolkit that lets AI agents flex their verified-human backing. It marries those proof-of-personhood creds with the x402 micropayments protocol from Coinbase and Cloudflare, because what's an AI agent without a way to request gas fees?

Even Ethereum's own Vitalik Buterin has weighed in, suggesting that crypto magic—specifically zero-knowledge proofs—could be the key to making online systems less trust-based and more verify-based. It's the ultimate "don't trust, verify" applied to the entire internet.

All this brainstorming is happening while regulators in Washington engage in their own form of soul-searching, which mostly involves writing frameworks. On March 20, the current administration dropped a national AI plan calling for a unified federal strategy, warning that a messy patchwork of state laws could kill innovation deader than a bear market.