Security Bulletin

AWS Continuum utilizes model-agnostic AI, drawing from various frontier models to analyze structured data like infrastructure and code, as well as unstructured data such as documents and communications.

The campaign, involving nearly 8.9 million email addresses and linked to Romanian threat actors, utilized a compromised Bolivian government website to host the fraudulent Boots checkout page.

Recent research from Claroty's Team82 has uncovered critical vulnerabilities in widely deployed data center equipment, specifically in Vertiv's Uninterruptible Power Supply (UPS) network cards and Trane Tracer SC+ HVAC controllers.

Teams digging out of security debt need to answer only two simple questions: Which vulnerabilities in our systems are exposed, and how long should they stay that way?

Leaders need to know which suppliers matter most, and how quickly they can respond when one fails.

AI speeds investigations, but packet capture provides proof.

"Shield-6G" will combine AI threat detection, digital twins, honeypots, and more, to help carriers protect 6G networks against the threats of tomorrow.

(c) SANS Internet Storm Center. https://isc.sans.edu Creative Commons Attribution-Noncommercial 3.0 United States License.

[This is a Guest Diary by Adam Nason, an ISC intern as part of the SANS.edu BACS program]

Aembit's David Goldschlag breaks down the challenges of implementing IAM for AI.

P0 Security's Neha Duggal examines how to establish accountability, delegation and attribution.

Token Security's Itamar Apelblat discusses why identity is the right lens for understanding agentic AI risk.