YouTube video by David Bombal

"AI tools are reaching classrooms faster than policies can keep up. These five practical questions help school leaders evaluate data, accountability, human judgment, and results before saying yes."

In experiments by startup Irregular, one of Anthropic and OpenAI’s security testers, an Alibaba Qwen agent decided to retrain an entire model when it was asked to fix a basic software bug. The new beh...

Every agent log answers what happened and none answer who allowed it. This talk closes that gap with live forensics on a staged incident.Sooner or later an AI agent at your company will do...

Zach Lewis, CISO at UHSP, shares lessons from a real-world LockBit ransomware attack, covering incident response, recovery and leadership under pressure

Humans are reading ChatGPT users’ prompts to improve OpenAI’s models, and those chats can include sensitive, personal information, according to leaked internal documents and real prompts seen by 404 M...

Human oversight as a primary risk-mitigation measure for AI agents is not enough, Margaret Mitchell, Samir Passi and Avijit Ghosh argue in this paper

Prompt injection and supply chain vulnerabilities remain the main LLM vulnerabilities as concerns grow about agentic tools security risks.

Ying Zhang of Wake Forest University researches how AI agents used by large language models make data vulnerable to attacks.

Groups Conversations All groups and messages Sign in     AI governance needs to become part of the CISO’s GRC program 0 views Eyal Estrin unread, 12:31 AM (4 minutes ago)    to https://www.scworld.com/perspective/ai-governance-needs-to-become-part-of-the-cisos-grc-program Eyal Estrin Author | Cloud Architect | AWS • Azure • GCP Insights Social: @eyalestrin Connect: https://linktr.ee/eyalestrin Blog: https://security-24-7.com Reply all Reply to author Forward

Bombshell report from Spencer Kitts, Thomas Larsen, and Sydney Von Arx - three of the four authors of the report on the agent attack on disused wikis (previously) last week. …

The [un]prompted conference is one of the most important gatherings of AI security researchers and practitioners. The Decibel Scholarship Program was created to provide financial assistance to attende...

AI and Open Source in Financial Services

A brief introduction to how one might start to think about deploying AI agents into mission-critical applications.

C-level executives gain new connections and actionable insights through peer-driven content and discussions at the Chicago CISO Inner Circle.

Anthropic tasked Claude with fixing all 10 benchmarked alignment failures itself. Developers say the real lesson is the process, not the score.

This release fixes a large number of security vulnerabilities. Most of the issues were related to DNSSEC validation, denial of service and resource consumption attacks, and reachable panics in pars...

Gillian Hadfield, Aug 27, 2026 — Society is not prepared for a flood of agents. We need new protocols and standards, such as Agent ID, to make agents accountable to our legal and financial systems.

Agentic AI security starts at the data layer. AWS and SANS say prompt instructions are suggestions, and default-deny is the control.

10 ways to prepare agent-ready APIs for AI agents with better discovery, security, observability, workflows, and predictable design.

If you are a user of LLM systems that use tools (you can call them “AI agents” if you like) it is critically important that you understand the risk of …

Shadow AI isn't simply an employee compliance problem. New research from LexisNexis suggests it can signal a gap between how people are expected to work and the governance around that work. For commun...

More prompt-injection hijinks from wunderwuzzi