The piece surveys developments in blockchain-enabled AI, where models run in verifiable, on-chain or hybrid on/off-chain systems to support permissionless, verifiable AI workflows even amid regulatory pressures.

It argues that frontier AI labs and top executives are pushing for regulation as a strategic moat, a move reminiscent of past proposals that aimed to shield incumbents from new entrants.

The article closes by underscoring a coming era as a pivotal fight between open, verifiable AI and regulatory control, with a teaser for future updates.

Gas Killerz is described as another on-chain approach—storing weights on-chain while performing inference off-chain with a signed, verifiable result to address speed and cost.

The broader theme is that AI governance battles could determine whether AI development remains open and verifiable or becomes confined behind permissioned, regulated barriers.

Regulatory pressure is framed as a tool to slow or block competition from open-weight AI models and foreign labs, potentially preserving incumbents like Anthropic and OpenAI.

Examples such as Fogo and Baranos AI run large language models with on-chain proofs of weights and off-chain computation to achieve ultra-low latency and verifiable outputs on a blockchain.

These on-chain AI efforts illustrate a broader move toward a permissionless, autonomous blockchain economy that could operate largely independently of centralized regulation, contrasting with regulatory capture efforts.