WuBlockchain
Binance founder Changpeng Zhao said on the PBD Podcast that everyone will eventually run thousands of AI agents alongside their human workflows, and that user metrics in the AI era will need to combine both. The remarks, dated May 12, 2026, frame a near-term shift in how crypto platforms measure reach, automate operations, and price infrastructure.
Why it matters
CZ disclosed that Binance currently pays roughly $10 million per month in AI-related fees — a line item that, if accurate, would rank among the largest disclosed AI cost bases in crypto. He said he has pushed the team to deepen AI use in risk monitoring, risk control, and compliance, but flagged that today's models still lack the domain-specific training and data needed for those functions. The gap between intent and deployability is the actual story: the spend is real, the workloads aren't yet plug-and-play.
Market impact
For exchanges, agentic AI is shifting from a productivity tool to a structural cost line and a compliance frontier. Risk and AML are the exact surfaces where errors carry regulatory teeth, which is why CZ's caveat about model readiness matters more than the headline number. Watch for peer disclosures and AI-vendor partnerships — the next leg is who supplies the trained, audit-friendly models exchanges will actually be allowed to put into production.
Source: [The $110 Billion Dollar Man - Binance Founder Opens Up | PBD #797 — YouTube](https://www.youtube.com/watch?v=Nu2qmRxwH4M)
Frequently asked questions
-
What did CZ say about AI agents on the PBD Podcast?
On May 12, 2026, CZ said everyone should eventually run thousands of AI agents alongside their human workflows, and that user metrics in the AI era will need to combine both humans and agents.
-
How much does Binance spend on AI each month?
CZ disclosed that Binance currently pays roughly $10 million per month in AI-related fees, one of the largest such cost bases publicly cited by a crypto exchange.
-
Where does Binance want to deploy AI next?
CZ said he has pushed the team to deepen AI use in risk monitoring, risk control, and compliance — but warned that current models lack the domain-specific training and data those functions require.
-
Why is AI in compliance a hard problem for exchanges?
Risk monitoring and AML are the surfaces where model errors carry direct regulatory consequences, which is why CZ's caveat about model readiness is a constraint on deployment rather than a productivity comment.
-
What should the market watch after CZ's AI comments?
Peer exchange disclosures of AI spend and partnerships with vendors supplying trained, audit-friendly risk and compliance models — the structural shift is in the cost line and the vendor stack, not the headline use case.
CoinTelegraph
WatcherGuru
CoinDesk
CryptoSlate
CZ Binance