AI-SDLC

Join us for the AISec Eng Poland Meetups!

Agenda:

17h - Talk 0: BSides Kraków Community updates - by Cássio Pereira 17h15 - Talk 1: Findings triage and fix with AI agents - Rafaela Durlo 18h - Talk 2: Threat Modeling with AI - Cássio Pereira

Overview for the talk: Every AI coding agent you run is an external entity with a shell, your credentials, and permission to act on your systems. When it works, that is the point. When it goes wrong, whether from a prompt injection buried in a web page it read or an accidental meltdown where the model does something destructive with no attacker involved at all, you have handed a potentially unpredictable process a terminal session with your access. The fix is not a new magic AI powered layer, it is the security we already know. Defense in depth still wins. Isolation and sandboxing to contain the blast radius, least privilege on tools, supply chain discipline, and above all monitoring and logging so you can see what the agent actually did. We will walk the layers with controls for real agent harnesses, then turn to the investigation side. When an incident happens, you reconstruct the agent's session from the artifacts it leaves behind and you will leave with a practical maturity model and the reassurance that your cyber security existing instincts, applied to a new kind of endpoint, are exactly right.

Whether you are a seasoned professional in the security industry or simply interested in learning more about hacking and technology, this meetup is the perfect opportunity to network with like-minded individuals and expand your knowledge. Don't miss out on this chance to enhance your understanding of cybersecurity and contribute to a safer digital environment for all.

And of course, we invite you for some drinks and appetizers to socialize AI-S