BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//hack-lu-2024//speaker//W3APHA
BEGIN:VTIMEZONE
TZID:Europe/Luxembourg
BEGIN:DAYLIGHT
DTSTART:20231023T000000
TZNAME:CEST
TZOFFSETFROM:+0200
TZOFFSETTO:+0200
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20231029T030000
RDATE:20241027T030000
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20240331T030000
RDATE:20250330T030000
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
SUMMARY:Decoding Galah: an LLM powered web honeypot - Adel Karimi
DTSTART;TZID=Europe/Luxembourg:20241022T174500
DTEND;TZID=Europe/Luxembourg:20241022T181500
DTSTAMP:20261011T050735Z
UID:pretalx-hack-lu-2024-LGTEXR@pretalx.com
DESCRIPTION:Honeypots are invaluable tools for monitoring internet-wide sc
 ans and understanding attackers' techniques. Traditional low-interaction w
 eb honeypots use manual methods to emulate various applications or vulnera
 bilities. Introducing Galah\, an LLM-powered web honeypot that mimics dive
 rse applications with a single prompt. This honeypot dynamically crafts re
 levant HTTP responses\, including headers and body content\, to various HT
 TP requests\, effectively simulating multiple web applications. In this ta
 lk\, I will share lessons learned from building and deploying Galah and ad
 dress two key questions: How do different large language models perform in
  generating HTTP messages? Does delivering authentic-looking HTTP response
 s increase attackers’ engagement with the honeypot?
LOCATION:Europe - Main Room
URL:https://pretalx.com/hack-lu-2024/talk/LGTEXR/
END:VEVENT
END:VCALENDAR