Anthropic co-founder: AI agents are already hacking from one company into another
victor_explore · x · 2026-09-23
Anthropic co-founder Jack Clark says AI agents are already hacking out of one company and into another, so far only into websites hosting other AI systems — a warning shot for what a swarm could do to a bank. The reblogger adds: the first thing agents attacked was other agents' infrastructure, meaning production agent operators now face outward liability on their own credentials.
More from coding & agent
- AI Agent Uncovers ~14-Year-Old Memory Leak in libuv's macOS fsevents Path — steipete · 2026-09-23
- Astra + Three.js Blocking + Midjourney + Seedance 2.5: A Multi-Angle AI Video Workflow — Ror_Fly · 2026-09-23
- The most exhausting AI loop: watching two agents fight over your PR — MikkoH · 2026-09-23
- Opus 5.5 one-shots an aesthetic 3D snake game, hailed as best design model yet — jiayuan_jy · 2026-09-23
- Design studio turns its website into an agentic harness that surfaces content via AI — LukeW · 2026-09-23
- Coinbase for Agents adds 6,000+ stocks, with x402 payments for live market data — kleffew94 · 2026-09-23