Ben Kliger

Zenity has been recognized as a Representative Vendor in @Gartner_inc’s Market Guide for Guardian Agents. 🛡️ As AI agents scale across SaaS, cloud, and endpoints, security must move beyond logging and prompt filtering. 🔒 🔎 Visibility 🧩 Governance 📊 Runtime detection ⛔ Inline prevention Read the report here: eu1.hubs.ly/H0s9ZGZ0 Learn more: eu1.hubs.ly/H0s9Ydp0 #AISecurity #AgenticAI #AIGovernance

we hijacked perplexity comet by sending a weaponized calendar invite then used it to takeover victim's 1p account and exfil their local files call it pleasefix. like clickfix, but instead of social eng'ing a human you just ask their ai real nicely incredible work by @StAJect0r

raptor found the @cline npm compromise payload

We are feeling it every day and are seeing the significant impact in every interaction we're having. Exciting times that's for sure! And MOST IMPORTANTLY- no need to panic. Learn-understand-adapt-repeat. Now 100x faster.

@Azure @zenitysec Super cool and proud to see this partnership come to life! #YallaZenity

@zenitysec Love seeing security front and center! Zenity + Foundry = visibility, hard boundaries, and runtime protection for agent fleets. 🙌 #MSIgnite

#MSIgnite delivered a big moment today. 🤩 During the AI Operations session for @Microsoft Foundry, Sarah Bird, Microsoft’s Chief Product Officer of Responsible AI, highlighted Zenity as a key security partner bringing inline prevention directly into the Foundry Control Plane. 🛡️ This comes just one day after we announced our expanded integrations with Microsoft Foundry and Copilot Studio. The timing could not be better. As enterprises move from experiments to large-scale agent fleets, the need for real-time enforcement and trustworthy controls becomes mission-critical. ⚡ Zenity provides that confidence by giving security teams visibility, hard boundaries, and runtime protection for agents built across the Microsoft ecosystem. 👁️ Learn more about the session here ➡️ ignite.microsoft.com/en-US/sessions… #MSIgnite #Foundry #CopilotStudio #AI

Last week I got to speak at Zenity's AI Agent Security Summit in SF I showed how I got a fintech agent to fall victim to goal manipulation 👿👇

Had a fantastic time presenting at the second AI Agent Security Summit! This time in SF. Great talks, great people, and great conversations. Big thanks to @zenitysec for hosting an awesome event! 🔥 And thx @mbrg0 for taking this picture.

@Nora_Carter_ Hard to pick just one :)

@benkliger Sounds like it’ll be epic 🔥 what session are you most excited about?

the #AIAgent security summit in SF next week is going to be legen.....dary! #nerdsonly #zenitylabs #agentflayer

@wunderwuzzi23 brilliant work 💪🏻 embracethered.com/blog/posts/202…

Devious new attack class defined by @wunderwuzzi23: "Cross-Agent Privilege Escalation" If someone has both Claude Code and GitHub Copilot on their machine you can prompt inject one and have it edit the config for the other in order to further escalate your attack's privileges!

The @zenitysec 2nd AI Agent Security Summit will take place on October 8th in San Francisco. Industry leaders - both makers and breakers - will be on stage. Secure your place! zenity.io/resources/even…

sign up: tinyurl.com/ZenityLabsSFSu…

we talked ai agents, enterprises racing to adopt immature tech, and hard boundaries we discovered while poking at flaship ai agents products resilientcyber.io/p/resilient-cy…

cfp for the ai agent security summit in sf is closing tmrw! its a really cool event. intimate room w/ ai security experts and enterprise execs. we want your hacking stories security program try to mng the problem off-the-record talks are welcome sessionize.com/zenity-ai-agen…

writeup for the salesforce einstein 0click hijack by @tamirishaysh labs.zenity.io/p/prompt-mines…

Common vulnerability i've seen across multiple vendors over the last few months: 👉 Agents that can modify their own configuration settings and security controls This is a recipe for RCE - and AI that escapes. More examples soon! 🙂

Full details of CVE-2025-53773 (RCE via Copilot in VSCode, Visual Studio, and other IDEs..) thanks to @gitlab and in particular @joernchen for the collaboration.

💥 Remote Code Execution in GitHub Copilot (CVE-2025-53773) 👉 Prompt injection exploit writes to Copilot config file and puts it into YOLO mode, then we get immediate RCE 🔥 Bypasses all user approvals 🛡️ Patch is out today. Update before someone else does it for you