BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//bsidesluxembourg-2026//speaker//E89977
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-bsidesluxembourg-2026-QW3PJK@pretalx.com
DTSTART;TZID=CET:20260507T144000
DTEND;TZID=CET:20260507T152000
DESCRIPTION:With the use of AI agents catching wind across the offensive se
 curity space\, from phishing\, to fuzzing and penetration testing\, it was
  inevitable that malware would follow suit. While most discussions focus o
 n using AI to generate malicious payloads at the malware’s runtime\, or 
 "vibe coding" it\, we went a step further: we built a system where AI is t
 he sole participant in the malware creation process itself.\nWe will begin
  by talking about how we got to this point\, what sparked the idea\, and j
 ump into comparing different models - showing which gave the best code\, w
 hich was most evasive\, which prompts worked the best\, and what we used i
 n the agent.\nWe will then dig into the generation process itself – we w
 ill show the challenges with earlier implementations\, and how we solve th
 em\, how to build the workflow to maximize the malware’s capability and 
 randomization\, and even how it managed to break signatures.\nWe will fini
 sh by showing how the resulting malware is performing\, comparing differen
 t samples\, and showing how each sample defeated several static malware an
 alyzers\, as well as talk about what's next for this agent\, and what's ne
 xt in the domain of AI-generated malware.
DTSTAMP:20260412T011329Z
LOCATION:IFEN room 2\, Workshops and AI Security Village  (Building D)
SUMMARY:The Agents of Chaos: AI Driven Malware Generation - Arad Donenfeld
URL:https://pretalx.com/bsidesluxembourg-2026/talk/QW3PJK/
END:VEVENT
END:VCALENDAR