BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
DTSTART:20250309T030000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:EDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251102T010000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:EST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260305T040936Z
UID:A0AFE771-192B-44C8-A0AA-46B3592C1A04
DTSTART;TZID=America/New_York:20250626T200000
DTEND;TZID=America/New_York:20250626T210000
DESCRIPTION:Generic Large Language Models (GLLMs) are continuously being re
 leased with increased size and capabilities\, promoting the abilities of t
 hese tools as universal problem solvers. While the reliability of GLLMs' r
 esponses is questionable in many situations\, these are augmented/ retrofi
 tted with external resources for different applications including cybersec
 urity.\n\nThe talk will discuss major security concerns of these pre-train
 ed models: first GLLMs are prone to adversarial manipulation such as model
  poisoning\, reverse engineering and side-channel cyberattacks. Second\, t
 he security issues related to LLM-generated codes using open-source librar
 ies/codelets for software development can involve software supply chain at
 tacks. These may result in information disclosure\, access to restricted r
 esources\, privilege escalation\, and complete system takeover.\n\nThis ta
 lk will also cover the benefits and risks of using GLLMs in cybersecurity\
 , particularly\, in malware detection\, log analysis\, intrusion detection
 \, etc. I will highlight the need for diverse AI approaches (non-LLM-based
  smaller models) trained with application-specific curated data\, fine-tun
 ed for well-tested security functionalities in identifying and mitigating 
 emerging cyber threats including zero-day attacks.\n\nSpeaker(s): Prof. Di
 pankar Dasgupta\n\nAgenda: \n20:00 Hours - START of Distinguished Lecture\
 n21:00 Hours - END of Distinguished Lecture\n21:00 Hours - Start of Q&A\, 
 Discussions\, Thoughts\, etc\n\nALL times are in EDT/EST format\n\nVirtual
 : https://events.vtools.ieee.org/m/486071
LOCATION:Virtual: https://events.vtools.ieee.org/m/486071
ORGANIZER:rkrish@ieee.org
SEQUENCE:31
SUMMARY:Generic LLMs in Cybersecurity
URL;VALUE=URI:https://events.vtools.ieee.org/m/486071
X-ALT-DESC:Description: <br /><p>Generic Large Language Models (GLLMs) are 
 continuously being released with increased size and capabilities\, promoti
 ng the abilities of these tools as universal problem solvers.&nbsp\; While
  the reliability of GLLMs' responses is questionable in many situations\, 
 these are augmented/ retrofitted with external resources for different app
 lications including cybersecurity.</p>\n<p>The talk will discuss major sec
 urity concerns of these pre-trained models: first GLLMs are prone to adver
 sarial manipulation such as model poisoning\, reverse engineering and side
 -channel cyberattacks. Second\, the security issues related to LLM-generat
 ed codes using open-source libraries/codelets for software development can
  involve software supply chain attacks.&nbsp\;These may result in informat
 ion disclosure\, access to restricted resources\, privilege escalation\, a
 nd complete system takeover.</p>\n<p>This talk will also cover the benefit
 s and risks of using GLLMs in cybersecurity\, particularly\, in malware de
 tection\, log analysis\, intrusion detection\, etc. I will highlight the n
 eed for diverse AI approaches (non-LLM-based smaller models) trained with 
 application-specific curated data\, fine-tuned for well-tested security fu
 nctionalities in identifying and mitigating emerging cyber threats includi
 ng zero-day attacks.</p><br /><br />Agenda: <br /><p>20:00 Hours - START o
 f Distinguished Lecture<br>21:00 Hours - END of Distinguished Lecture<br>2
 1:00 Hours - Start of Q&amp\;A\, Discussions\, Thoughts\, etc<br><br>ALL t
 imes are in EDT/EST format</p>
END:VEVENT
END:VCALENDAR