Security & Threat Intelligence · 31.07.2026, 12:48 UTC
Enhancing AI security through global AI red teaming
| Schweregrad | info |
|---|---|
| Kategorie | Security & Threat Intelligence |
| Quelle | Microsoft Security Blog ↗ |
| Veröffentlicht | 31.07.2026 UTC |
Sicherheitsmeldung mit Schweregrad noch nicht bewertet. Technische Details im Tab „Originaltext“; empfohlene Schritte in der Checkliste.
In this article
Building a global allianceWhat the research focuses onWhy this matters
Most AI safety testing still happens inside the walls of individual organizations. That has resulted in a fundamental disconnect: many of the highest-risk failure modes in modern AI systems require deep domain expertise, multilingual context, or regional understanding that no single internal team can fully replicate on its own.
As frontier models become more capable, the attack surface expands with them. AI red teaming is no longer just about prompt injection or content safety edge cases. It increasingly involves security operations, misuse scenarios, multilingual harms, alignment failures, and domain-specific abuse patterns that can vary significantly across geographies and languages.
Microsoft’s AI Red Team has observed that meaningful testing of advanced AI systems- and models similarly requires broader participation from researchers and practitioners who operate outside traditional corporate security boundaries. To address that gap, today we are announcing the External Red Team Alliance (EXTRA), a formalized global extension of Microsoft’s AI Red Team designed to support and encourage external expertise to advance AI safety and security testing. We are proud to share we are funding the development of new AI safety assessments on six continents through unrestricted gifts.
Building a global alliance
EXTRA is a two-part initiative focused on expanding AI safety research and strengthening external collaboration.
The first component supports a global …