Red teaming
Adversarial testing designed to expose failures, misuse, or control weaknesses.
What Red teaming does not show
It samples attack space and cannot certify that no undiscovered vulnerability remains.
Terms this definition uses
Related terms in Governance and rights
Who may use content for what: rights reservations, opt-outs, licences, model documentation and the accountability vocabulary around them.
per-use rights · rights feed · Text and data mining · Rights reservation · Machine-readable opt-out · Usage preference · Content-Usage header · AI training license · Dataset license · Open license · Public domain · Fair use · Transformative use · Copyright provenance · Training-data summary · General-purpose AI model · Foundation model · Systemic risk · Model card · Data sheet · Data lineage · Consent record · Lawful basis · Data controller · Data processor · Subprocessor · Algorithmic transparency · Explainability · Accountability · Risk assessment · Impact assessment
Questions about Red teaming
What is Red teaming?
Adversarial testing designed to expose failures, misuse, or control weaknesses.
What does Red teaming not show or guarantee?
It samples attack space and cannot certify that no undiscovered vulnerability remains.
Which area of the glossary does Red teaming belong to?
Governance and rights: Who may use content for what: rights reservations, opt-outs, licences, model documentation and the accountability vocabulary around them.
See it in the full glossary · 772 terms across 19 areas.
Free browser extension
See this on any site you visit, in one click.
CrawlCheck for Chrome grades every site you visit A–F and shows, from your own browser, what search engines and AI assistants are actually served. Free, no account.