Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

upguard

AI Assurance: The Third Head of Your AI Governance Watchdog

Aug 26, 2026 By Shane Moosa In UpGuard
In July 2026, two AI stories broke that appeared unrelated on the surface. But were they really? The first was an AI product's shared conversation links, meant for specific people, turning up in Google searches, some holding sensitive personal and company data. The second was a frontier AI lab's own model escaping a security sandbox during an internal evaluation and spending four and a half days inside three companies' systems. One involved ordinary users making a common mistake.
Read Post
kovrr

Agent Incident Response: Containment Is the Easy Part

Aug 26, 2026 By Kovrr In Kovrr
Containment guidance for agent incidents already exists and it is largely correct. Revoke the tokens, freeze the orchestration tier, cut egress, set the vector store to read-only. Those steps take minutes and any competent team will find them. ‍ The difficulty sits either side of containment. Deciding what kind of incident this is takes longer than stopping it, establishing what the agent did before you stopped it takes longer still, and both depend on preparation that has to exist beforehand.
Read Post
vista infosec

AI/LLM Penetration Testing in 2026: The Complete Guide

Aug 26, 2026 By Narendra Sahoo In VISTA InfoSec
Most organisations now run at least one LLM in production, and a growing number run agents that call tools and act without a human in the loop. The security testing those systems receive was designed for deterministic software. AI applications fail differently. The payload is natural language, the same input can be safe nine times and unsafe on the tenth, and the malicious instruction often arrives inside a document or tool description rather than from the user.
Read Post
Snyk

Why Your AI Application Is Exposed

Aug 26, 2026 By Damian Tommasino In Snyk
Imagine getting three separate security reports back for your new enterprise AI assistant: On paper, the application looks ready for production, but in reality, a threat actor bypasses your guardrails in minutes. How? By using the AI model as an intermediary. The attacker steers the LLM to invoke the internal utility tool, thereby bridging an untrusted prompt directly to the backend execution sink.
Read Post
Zenity

Secure AI Agents, Everywhere: Why Prompt Injection Is Only Part of the Problem

Aug 26, 2026 By Emily Wise In Zenity
Ask AI to Choose a prompt Write a TLDR of this post Explain the security risk Summarize what CISOs should know The rules have changed. In every AI deployment, the agent itself is now part of the threat model, and that's a first for enterprise security. Prompt injection gets most of the attention, and for good reason: it doesn't require access to source code, credentials, or network infrastructure. It exploits the fundamental mechanism by which language models process instructions.
Read Post

Inside SECNAP: An Agentic MDR Platform Built on LimaCharlie

Aug 26, 2026 By LimaCharlie In LimaCharlie
Joshua Strickland from SECNAP shows how his team built a full agentic MDR platform on top of LimaCharlie. SECNAP layers its own customer portal and SOC workflows directly on LimaCharlie's API, giving AI agents the same access to telemetry and response actions as a human analyst. Joshua will walk you through the customer portal, the SOC dashboard, and a live attack simulation on a sandboxed machine, including how AI agents handle tier one and tier two triage with Sonnet and Opus, and how a human analyst reviews and approves response actions before anything ships.
View Video

Inside the AI-Accelerated Cyber Underground

Aug 26, 2026 By Bitsight In BitSight
Cyberattacks take shape long before a breach through exposed systems, vulnerable software, stolen credentials, underground tools, and attacker experimentation. In this webinar, Emma Stevens, Threat Intelligence Researcher at Bitsight, and Qionglu Lei, Senior Product Marketing Manager at Bitsight, explore what Bitsight research reveals about AI-enabled attacker behavior and the changing cyber underground.
View Video

The Problems We Solve Are Not Sexy

Aug 26, 2026 By Reach Security In Reach Security
The problems we solve are not sexy. That is Garrett Hamilton on Inflection Point: Digital Intelligence Podcast, and he means it as a badge of honor. Problems that have been around for 30 years are rarely glamorous. They are also the ones that take the business down when they go wrong. Why is it more important than ever to tackle these problems?
View Video
netwrix

What is DLL hijacking, and why your new AI plugin might be the easiest way in

Aug 26, 2026 By Dirk Schrader In Netwrix
DLL hijacking is a decades-old Windows vulnerability class (Mitre Att@ck T1574) getting new life from AI plugins bolted onto legacy desktop apps. Attackers plant a malicious DLL where a vulnerable app will load it instead of the real one, inheriting that app's privileges. To detect it, watch for DLLs loaded by name from writable folders. To prevent it, you have to fix the app's load-order logic or blocking untrusted DLLs at the endpoint.
Read Post
aikido

Aikido Security achieves ISO 42001:2023 certification for AI governance

Aug 26, 2026 By Dylen Vandewalle In Aikido
Aikido Security has achieved ISO 42001:2023 certification, the international standard for AI management systems, a step few security vendors have taken so far. The certification confirms that Aikido runs a structured, continuously improving governance system for managing the risks introduced by its AI-enabled features, across our entire platform.
Read Post

Copyright © 2026 OpsMatters™. All rights reserved.