TAICO April 2026 Meetup!
Details
The TAICO team is happy to announce our next meetup at the Adaptavist office in Toronto. Much thanks to Adaptavist for hosting!
Wayfinding: How to Get to the Event
- Please be careful when using maps features on phones, some people have been redirected to a completely wrong address. The event is near Winners on Front St, which is just south, across the street, from Berczy Park (where the dog fountain is) so if you are not near that area, you are in the wrong place.
Registration and Event Details
Our goal is to bring artificial intelligence and cybersecurity together. To do that, we need to explore what's happening in AI and what's happening in cybersecurity, and where the two intersect and collide. We're also working to understand how we solve problems in these areas, what that looks like, and how Canada can and will participate.
With that in mind, we're pleased to announce our speakers and the agenda for our next meeting!
Sponsors
And thanks to our host, Adaptavist!
Notes
- Unfortunately, we won't have pizza at this meetup due to a gap in sponsorship. We're actively working on securing sponsors for future events.
Agenda
Doors Open: 5:30PM
Event Start: 6:00PM
- 👋 Welcome and introductions
- 🙏 Thank you to our amazing hosts and sponsors! 🎉
- 🎙️ Speakers
🎤 Speaker #1: Dmitrii Strizhkov
Subject: Penetration Testing Automation with MCP
Abstract:
This presentation talks about my journey as the penetration testing practice lead, covering how I was able to increase efficiency and speed up the execution of my team's projects. I will discuss what MCP is, introduce how it can help in reconnaissance and enumeration phases, and analyze the results. I will also demo a solution that everyone can implement using the n8n low-code platform. Additionally, I will discuss and show why agents can help us, pentesters, in writing the reports (do the boring and tedious stuff). In the end, I will delve into why several precautionary measures should be undertaken before sending sensitive client data to commercial LLMs and suggest several solutions based on my experience. And as the final point, I would want to encourage my colleagues to experiment with and implement their own MCP-based solutions to speed up and make their engagements more effective.
About Dmitrii:
Dmitrii is the head of offensive security/penetration testing and practice lead at Parabellyx Cybersecurity.
🎤 Speaker #2: David Ahmann
Title: Stop Is Not a Safety Control: What AI Agent Security Looks Like at the Runtime Boundary
Abstract:
A lot of AI security discussion still focuses on prompts, policies, and tool declarations. Our research at (CAISI) has been looking at the harder question: what actually changes at execution time when an agent is told to stop, requires approval, or reaches a risky tool boundary? I’d share a practical, evidence-based talk on where AI agent control really holds or fails, drawing from our OpenClaw case-study work and our public research on agent governance, evidence posture, and approval gaps. The goal would be a practitioner-focused session for security leaders, builders, and evaluators trying to separate demo confidence from real control.
⚡ Lightning Talks and Demos
Lightning Talks - 5 to 15 minutes long
- Curtis Collicutt - A demo of a new programming language: Vary - A typed language design to find "lies" in generated code
- Janhavi Doshi - "Data, LLMs and the Law - What Every Builder in Canada Needs to Know Right Now"
- You? - Please reach out if you'd like to do a lightning talk or demo - anyone is welcome to demo something they are working one
Please reach out to us if you'd like to present at the meetup. We are looking for people to talk about what they are working on, what they are building and learning, and are open to any level of experience and technical depth. Whether you are a beginner or an expert, we want to hear from you! We're all just out here building and learning.
👋 See You There!
Thanks, and we look forward to seeing you at the meetup!