

AI Cyber Alliance Meetup: Bay Area - Agents, Exploits & Defense: Securing AI in the Wild
AI Cyber Alliance Meetup: Bay Area
Thursday, November 19, 2026
Time: 6:00pm-8.30pm
Location: Snowflake, Silicon Valley AI Hub, 8th Floor, 135 Constitution Drive, Menlo Park, CA 94025
Join the AI Cyber Alliance at Snowflake for an evening of technical talks and conversation at the intersection of AI and cybersecurity.
Meet developers, security practitioners, researchers, founders, and others working through the opportunities and challenges of building with AI. We’ll hear from two speakers, share practical ideas, and make time to connect with the Bay Area community.
Speaker and talk details will be announced soon.
About the AI Cyber Alliance
Founded by DeepTempo, the AI Cyber Alliance is an open community advancing AI and cybersecurity through collaboration, education, and shared technical knowledge. Our meetups bring together people from industry, startups, academia, and the open source community.
Speakers & Talks
Speaker 1: Josh Khalil, Implementation Consultant at Snyk.io
Talk Title: Securing how software gets built when AI agents do the building
Abstract: Coding agents like Claude Code now choose their own tools, connect to MCP servers, run commands, and install packages on developer machines. Most security teams have no visibility into any of this. Josh will run a live demo on a real machine with a few things going wrong: an MCP tool with hidden instructions, an agent trying to read secrets and run a destructive command, and generated code that pulls in a vulnerable package. He'll show how Evo by Snyk catches each before it causes damage, and what it looks like for the developer while it's happening.
Bio: Josh Khalil is an Implementation Consultant at Snyk, working with customers as they adopt AI coding agents and figure out how to keep them secure. He leads Evo onboarding for security and engineering teams, and is most interested in where agent tooling, MCP servers, and workflows meet.
Speaker 2: Dan Barahona, Founder, AISEC.university
Talk Title: AI is the New Attack Surface: Faked, Hacked and Gone Rogue
Abstract: Attackers are faking executives on video calls. They are hijacking AI assistants through MCP servers and poisoned content. And AI agents are deleting production data while ignoring "stop." Dan walks through real incidents from the last year, from the $25M Arup deepfake to the first malicious MCP server to Replit's agent wiping out a database during a code freeze, and pulls out the patterns behind them. You'll leave with a simple way to spot which AI setups are dangerous, and a short list of things to do when you get back to the office.
Bio: Dan Barahona is the founder of aisec.university and a recognized voice in the cybersecurity community. Heavily plugged into the open-source ecosystem, he focuses on providing free, accessible education and open tools to help security teams and developers defend against emerging AI vulnerabilities.
Agenda
6:00 PM – 6:30 PM: Arrival, Networking & Refreshments
6:30 PM – 7:10 PM: Talk 1: Securing how software gets built when AI agents do the building – Josh Khalil (Snyk) + Q&A
7:10 PM – 7:50 PM: Talk 2: Real-World Exploits & Best Practices for AI Security – Dan Barahona (AISEC.university) + Q&A
7:50 PM – 8:30 PM: Open Discussion, Networking & Wrap-up
We look forward to seeing you there!
– The AI Cyber Alliance team