Security· ★★★· neutral·

CertiK Urges Web3 Firms to Treat AI Agents as Workforce Members

  • —CertiK calls for explicit roles, bounded authority and human accountability for AI agents
  • —AML penalties exceeded $900 million in H1 2025
  • —Report urges mandatory auditing of AI agents' reasoning logs
  • —CertiK backs real-time onchain circuit breakers to halt exploits
Why it matters: If Web3 protocols adopt AI log audits and automated circuit breakers, exploit losses could shrink and regulators may set new security standards for the sector.
Source: Bitcoin.com
More on this
Regulation · Oct 2
HYPE Holds Near $90 as Hyperliquid Urges EU to Treat Perps as Derivatives
 Security · yesterday
CertiK: AI agents now trace stolen crypto, but humans still own the mistakes
 Security · Oct 5
CertiK: Agentic AI Set to Reshape Crypto Cybersecurity, AML and Compliance