Monzy Merza on the Rise of Agentic SOC

CISOs are facing challenges in tracking their AI applications, let alone securing them. As interactions among agents increase, the visibility issue only grows more critical. Pentera’s approach offers a solution by integrating AI-driven attack emulations with clear guardrails, enabling security teams to act swiftly while maintaining oversight. This method allows organizations to assess their actual vulnerabilities in an ever-changing landscape. Pentera’s Aviv Cohen, Chief Marketing and Corporate Strategy Officer, highlighted the industry’s AI conundrum: "You want the honey, but you don't want the sting." Join Aviv as he explores a major transformation in cybersecurity: the emergence of AI-driven threats, alongside Jarrett Banks from Tech Edge. https://okt.to/LERmcX #AIAttacks #AdversarialAI CorpGov

The safest AI model Anthropic has ever built is also the most dangerous one you’ll deploy this year.    The Mythos release claims to change cybersecurity forever, but the reality is more about marketing than a shift in safety. While the headlines focus on zero-day vulnerabilities, they often ignore that these attacks require specific access and contexts that many secured infrastructures don't even allow.   Beyond the hype, the real danger is the looming "price spike." If you aren't prepared for token-based pricing and budget increases, your projects risk being throttled or failing entirely by 2027.   We’ve officially launched a new series called Hype Check in Tech on YouTube, where we cut through the noise and vendor spin to tell you what actually matters for IT and security teams. Give us a follow and let us know in the comments what topics you would like us to cover. Watch the full episode here: https://avpt.co/4wrjwsz #aigovernance #mythos #claudemythos 

Opinion: In Reality, Mythos is the Finest Cybersecurity News We’ve Ever Encountered #internet #cybersecurity #rswebsols https://ift.tt/3zCqRyQ Excited to share our latest analysis on the Mythos debate in cybersecurity. The piece examines Anthropic’s Mythos model, exploring whether automated vulnerability identification could level the playing field for defenders, and the broader policy and governance implications as nations and firms increasingly leverage AI for cyber defense and offense. It also discusses open-source security challenges, access disparities to AI tooling, and the potential for a paradigm shift in how we secure software before deployment. Read the full perspective and join the conversation on how AI-enabled vulnerability discovery could reshape our security landscape. https://ift.tt/3zCqRyQ

“Data islands” pose serious risks for SOC teams. When telemetry, alerts, identity data, and threat intel live in separate tools that don’t talk to each other, AI is forced to operate with blind spots. Andesite's Chief Product Officer, William MacMillan recently spoke at ADAPT’s Data & AI Edge Conference about the importance of connecting data islands. MacMillan explained how Andesite's technology collects and synthesizes data from disconnected sources so analysts can quickly and accurately assess security events - saving hours of time and drastically reducing risk. Check out this clip from his interview. #Cyber #CyberData #Cybersecurity #HumanAISOC #SOC #AIData

Our CEO, Dan Lahav, spoke at J.P. Morgan's Global Cyber Innovation Summit in New York City. As AI systems move into production, security teams need better ways to understand how they behave before attackers do. Dan's session focused on cybersecurity in the era of frontier AI, exploring how AI systems fail and how to build trust where traditional security tools fall short. A timely conversation for a fast-moving field.

𝐋𝐲𝐫𝐢𝐞.𝐚𝐢 𝐋𝐚𝐮𝐧𝐜𝐡𝐞𝐬 𝐑𝐞𝐚𝐥-𝐓𝐢𝐦𝐞 𝐙𝐞𝐫𝐨-𝐃𝐚𝐲 𝐓𝐫𝐚𝐜𝐤𝐢𝐧𝐠 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐀𝐠𝐞𝐧𝐭𝐢𝐜 𝐀𝐈 𝐄𝐫𝐚 Lyrie.ai has announced a new real-time zero-day tracking and disclosure system designed to help enterprises detect active exploit activity faster, alongside acceptance into Anthropic’s Cyber Verification Program and the launch of the open Agent Trust Protocol (ATP) to secure autonomous AI agent identities across enterprise environments. “The difference between a breach and a near-miss is usually measured in hours. We built Lyrie to be the system that finds the threat before it finds you,” said Guy Sheetrit, CEO and Founder of OTT Cybersecurity LLC. Read more: https://lnkd.in/d9JHbR8S #CyberSecurity #AI #ZeroDay #ThreatIntelligence #AgenticAI #AISecurity #EnterpriseSecurity #LyrieAI #CyberTech #IdentitySecurity

Modern attacks don’t start and stop at the endpoint. When attackers bypass controls, the hybrid network is where they move, pivot, and spread. That’s why Vectra AI’s Network Detection and Response focuses on detecting attacker behavior inside the network, where visibility is often lost. In this video, see how Vectra AI NDR helps security teams cut through alert noise and respond faster to real threats across modern environments. Learn more: https://lnkd.in/gbGNcNHY #CyberSecurity #NDR #AI #ThreatDetection #VectraAI

AI agents are creating a new governance problem for security teams. As they interact with data, tools, and business logic, security needs to answer complex questions in real time: what is the agent trying to do, what authority does it have, and did the user even request it? That is the focus of Unified Agentic Defense Platforms. On May 21st, I’ll be joining Lawrence Pingree, Head of Data Security and AI Research at Software Analyst Cyber Research, for a discussion on what this emerging category means for CISOs and security teams. May 21 | 11 a.m. ET / 8 a.m. PT Register here: https://lnkd.in/gpeAnh-y #CyberSecurity #AIAtWork #AIGovernance

The biggest cyber threats are often the ones you can’t see. Shadow IT and lateral movement allow attackers to stay hidden until the damage is done. See how vSOCBox uses AI/ML-powered detection, behavioral analytics, and automated response to uncover unauthorized assets, stop threat propagation, and reduce breach risk in real time. Read the full casestudy: https://zurl.co/FjqBS Book a Demo: https://zurl.co/AKtAc #CyberSecurity #ShadowIT #LateralMovement #AI #MachineLearning #SOC #ThreatDetection #vSOCBox

The first company to acquire Last Label Encryption will own the next decade of Cybersecurity. We’re seeing a massive $50M+ surge in "Human Risk" platforms this week, and the message is clear: the human layer is the new perimeter. But training people is only half the battle. The missing link? A technical fail-safe that makes human error irrelevant. Most encryption protects data in transit or at rest. LLE protects it at the final point of consumption—the "Last Label." When you combine AI human defense with LLE's mathematical certainty, you don't just reduce risk; you neutralize the threat entirely. In an era of deepfakes and automated vishing, the major player who moves first on LLE won't just be adding a feature—they’ll be defining the Sovereign Security standard for 2026 and beyond. Who do you think makes the move first? 🏗️🔐 #Cybersecurity #LLE #CISO #DeepfakeDefense #Innovation