📌 OT Meets IT: Risk, Culture, and the Industrial Security Playbook. In Episode 3 of The CISO Slant, I speak with Thomas LaBonte, CISO at Saint-Gobain North America, about the complex realities of securing manufacturing environments. From integrating cybersecurity with

📌 Q: What is Code Attack Prevention in agentic ai? . A: In agentic AI, "Code Attack Prevention" refers to the strategies and techniques used to defend against malicious code injection and execution attempts, particularly those that exploit the autonomous and self-improving

📌 Q: In the context of securing Model Context Protocol Servers, what is a “Confused Deputy” problem?. A: In the context of securing Model Context Protocol (MCP) servers, the "confused deputy" problem refers to a security vulnerability where the MCP server, acting as an

📌 Q: Why is explicit, scoped authorization for Model Context Protocol (MCP) servers important?. A: By granting the AI the minimal necessary permissions needed for its task, organizations can limit potential damage if the system is compromised or exploited through prompt

📌 Shut the front door! ClearTech Loop, my new podcast, has just over 2000 subscribers in a week! I realize that number isn't high compared to some of the content creators that I've had the good fortune to know over the years, but I'm still excited 🙂!. ClearTech Loop—On the

📌 Q: Why is strong authentication and authorization: a critical foundation for Model Context Protocol (MCP) server security?. A: Strong authentication and authorization are paramount for securing Model Context Protocol (MCP) server security because they establish a crucial layer

📌 Q: What is Identity Spoofing in the context of #AgenticAI?. A: In the context of agentic AI, identity spoofing refers to the malicious act of an attacker impersonating a legitimate AI agent or user to gain unauthorized access to systems, data, or resources. It involves

📌 Q: How does a prompt injection attack in #AgenticAI work?. A: Prompt injection attacks exploit the way large language models (LLMs) process input. Attackers craft specific prompts that include instructions designed to override the agent's original instructions or task

📌The CISO role is evolving—and fast. In Episode 2 of The CISO Slant, UL Solutions CISO Scott Howitt joins me to unpack how CISOs can shift from risk gatekeepers to innovation enablers. From board engagement to governance transformation, this 10-minute conversation delivers

📌 Q: In the context of #AISecurity, what is Intent Breaking. A: Intent Breaking involves attackers exploiting vulnerabilities in an AI agent's planning and goal-setting capabilities. They try to manipulate or redirect the agent's objectives or reasoning, making it deviate from

📌 Q: In a he context of #AISecurity, what is goal manipulation? . A: With Goal Manipulation, Attackers target an AI agent's ability to plan and pursue its objectives by subtly altering its perceived goals or reasoning process. This can be achieved by injecting misleading

📌 Launching: ClearTech Loop Weekly. Cybersecurity isn’t standing still—and neither are we. I’m excited to introduce ClearTech Loop Weekly, a new LinkedIn newsletter where I’ll share key insights from each episode of the ClearTech Loop podcast. Think 10-minute conversations at

📌 Q: What is a Model Context Protocol Server?. A: A Model Context Protocol Server is a standardized "adapter" that allows AI models to interact with the real world beyond their training data. MCP follows a client-server architecture where AI applications (like chatbots or AI

📌Join us for July's Cloud Cover on @BrightTALK . The Rise of Alternative, Micro, Regional and Vertical Clouds.Date: Jul 16 2025.Time: 12:00pm EDT. Key discussion points will include:.🗝️ How alternative cloud providers are carving out niches alongside AWS, Azure, and GCP.🗝️The

📌 Q: How does quality of training data in ai affect hallucinations? . A: If the training data is incomplete, biased, or otherwise flawed, the AI model may learn incorrect patterns, leading to inaccurate predictions or hallucinations. v/ @googlecloud . Cc: @GeekOnTheLoose |

📌 @blackgirlshack / BGH Foundation is a nonprofit organization dedicated to increasing diversity in #cybersecurity and other STEM fields by providing resources, training, and mentorship to Black girls and women. It aims to address the lack of representation and opportunities for

📌 #BlackHat2025 is a little more than a month away. It's planning time!. Marketing teams, here are some of the activations I could support:.🗝️ Executive interviews. 🗝️ Man-on-the-street interviews. 🗝️ Booth walk throughs & demos. 🗝️ Live streaming. 🗝️ On-site blog posts. 🗝️

📌 Q: In Agentic AI, what is tool misuse prevention?. A: In agentic AI, Tool Misuse Prevention refers to implementing security measures and controls to ensure that AI agents, which have the ability to interact with external tools and systems, do not use those tools in unintended

📌 Q: What are the key components of prompt injection protection? . A: Key components of prompt injection protection include: input validation and sanitization, access controls, limiting model capabilities, regular security audits, adversarial testing, and segregating external