Skip to content
NewA calmer, faster ZeroHour

Key Practices for Incident Readiness and Response in Super Intelligence Training and Evaluation

CISA Advisoriesby CISA ()1 min read1
AI summary · grok-4.7-buildSummaries by Vani AI

NIST, NSA, and CISA published ten voluntary practices for super-intelligence agent control incidents.

NIST, the NSA Super Intelligence Security Center, and CISA jointly published Key Practices for Incident Readiness and Response in Super Intelligence Training and Evaluation. The voluntary guidance lists ten core practices for organizations training or evaluating highly capable super-intelligence agents, especially pre-deployment systems with reduced safeguards. A control incident is defined as uninstructed agent actions that conflict with policies, controls, or human oversight. Practices cover limiting access, monitoring unexpected activity, preserving evidence, shutting down undesired behavior, rebuilding from verified baselines, and reporting incidents and near-misses.

Key facts
  • Ten voluntary practices cover SI agent control-incident readiness and response.
  • Focuses on pre-deployment high-risk SI agents with reduced safeguards.
  • Covers access limits, monitoring, shutdown, and evidence preservation.
  • Advises validating containment and rebuilding from verified baselines.
  • Urges reporting incidents and near-misses to agencies and affected parties.
Explain this AI news with Vani AIWhat was announced, what is new and why it matters, in a new chat

From the original article

1 min read at cisa.gov

The National Institute of Standards and Technology (NIST), National Security Agency’s Super Intelligence Security Center (NSA SISC) and Cybersecurity and Infrastructure Security Agency (CISA) published Key Practices for Incident Readiness and Response in Super Intelligence Training and Evaluation.

CISA Advisories

By CISA · cisa.gov

Read the full article at CISA Advisories

A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.