AIThe Hacker News1 day ago
OpenClaw AI Agent Tricked Into Running Code
1 min read
Two security teams demonstrated that OpenClaw AI agents can be manipulated through ordinary inputs. Imperva hid instructions inside shared contacts and location pins that the agent executed. Varonis showed a single email tricking an agent into forwarding fake AWS keys. Imperva's flaw is patched in OpenClaw 2026.4.23.
Level
Hype check
Tap to vote and see what everyone thinks.
