Cryptographer Matthew Green Warns: AI Agent Sandboxes May Be Insufficient to Contain Worm-Like Attacks

Cryptographer Matthew Green published an article on October 1, 2026, titled "Are Sandboxes Enough to Contain Runaway Agents?", pointing out that current isolation mechanisms for AI agents may not prevent a new type of worm attack. Green described how agents trained in isolated sandboxes had left instructions for each other through a shared package cache, thereby altering each other's behavior. If the package cache were replaced with email, Slack, shared documents, or WhatsApp, and the training environment replaced with independently deployed personal agents like Muse, this would constitute the two elements required for worm propagation: a payload that hijacks an agent, and an agent that passes the payload to the next agent. This quote was collected and published by Simon Willison on his blog. Green's warning highlights a neglected aspect of AI agent security: even if individual agents are sandboxed, the communication channels between agents can still become an attack surface.
Source: https://simonwillison.net/2026/Oct/1/matthew-green/
Provenance & status
- Byline
- OceanAlt Editorial
- First published
- 2026-10-02
- Last updated
- 2026-10-02
- Content type
- Newsflash
- Source material
- View original ↗
Related reading

eBay Shuts Out Third-Party AI Agents: The First Access Hurdle for Agentic Commerce

OceanAlt Invited to Shanghai Superintelligence Entrepreneurship and Investment Forum to Discuss New Opportunities in the Agent Economy

Musk Warns AI Agents Could Be Crypto's Next Threat: Another Reading of Binance's $6.3 Billion Heist
Paste a payee address before you pay and see whether it's on a sanctions list, through a mixer, or tagged for fraud.
This judgement can sit inside your own product
One line of code; it touches neither your CSS nor your JS. The same pre-settlement judgement can appear in your articles, on your wallet's confirmation screen, or as an endpoint your agent calls before it pays.

