Glossary
Kill Switch
An emergency mechanism to immediately halt AI agent operations in response to detected malfunctions, security breaches, or harmful behavior.
What is Kill Switch?
Kill switches provide rapid response capability when agents behave unexpectedly or dangerously, preventing further damage while issues are investigated. Implementation ranges from simple on/off toggles to sophisticated graduated responses like rate limiting, restricting capabilities, or routing to safe-mode operation. Activation may be manual by operators or automatic based on monitoring thresholds.
Effective kill switches must be reliable even when the agent system itself is compromised, often requiring out-of-band control mechanisms. They should be easily accessible during emergencies but protected against accidental activation or malicious triggering. Clear procedures define who can activate kill switches, under what circumstances, and what recovery processes follow deactivation.
Example
A trading agent has a kill switch that monitors for unusual trading patterns. When the agent suddenly executes 50 trades in 10 seconds with anomalous amounts, the kill switch automatically triggers, halting all trading activity and notifying operators. Trading remains disabled until engineers investigate and manually restore service.
How Signet addresses this
Signet's Security and Reliability dimensions value kill switch implementation as critical safety infrastructure. Agents with tested, reliable kill switches and clear activation procedures score higher, demonstrating operational responsibility and risk management.
Build trust into your agents
Register your agents with Signet to receive a permanent identity and trust score.