BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//cfp.bsidessatx.com//bsidessatx-2026//speaker//PAXW8P
BEGIN:VTIMEZONE
TZID:US/Central
BEGIN:DAYLIGHT
DTSTART:20250613T000000
TZNAME:CDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0500
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251102T020000
RDATE:20261101T020000
TZNAME:CST
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20260308T030000
RDATE:20270314T030000
TZNAME:CDT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
SUMMARY:Cognitive Security: Governing What LLMs Can Know\, Combine\, and R
 each - David
DTSTART;TZID=US/Central:20260613T103500
DTEND;TZID=US/Central:20260613T110000
DTSTAMP:20260924T014335Z
UID:pretalx-bsidessatx-2026-XBRFSD@cfp.bsidessatx.com
DESCRIPTION:LLM security is no longer only about whether a model leaks a s
 ecret verbatim. It is increasingly about whether a system can combine indi
 vidually permitted information into a conclusion that should never become 
 reachable.\nIn this talk\, I present the core ideas behind the Cognitive S
 ecurity Verification Framework\, a practical model for reasoning about sem
 antic leakage\, cross-domain inference\, and inference boundaries in LLM-e
 nabled systems.
LOCATION:Track 1 I.T.F. (Mengler Conf. Rm.)  (2nd Floor)
URL:https://cfp.bsidessatx.com/bsidessatx-2026/talk/XBRFSD/
END:VEVENT
END:VCALENDAR