BEGIN:VCALENDAR
VERSION:2.0
PRODID:Data::ICal 0.24
BEGIN:VEVENT
DESCRIPTION:   'Title: The Defender’s Dilemma: Releasing Dual-Use AI Mode
 ls as\n   Capabilities Climb\n   Tags: DEF CON Official Talk | AI Village\
 n   When: Friday\, Aug 7\, 13:30 - 14:59 PDT\n   Where: LVCCW Level 1 Hall
  3 1006 (Main Track 1) and DCTV-1 - [1]Map\n\n   Description:\n\n   Defend
 ers and attackers are locked in a renewed race where AI has\n   enabled ne
 w defenses and new attacks. AI models are gaining capability\n   at a rema
 rkable pace\, and dual-use systems\, powerful for defenders and\n   advers
 aries alike in domains like cybersecurity\, bring that trajectory\n   into
  sharp focus. As these models grow stronger\, developers face a\n   practi
 cal question: How can defenders and labs each play a part in\n   ensuring 
 that defenders win?\n\n   This panel takes up emerging best practices for 
 applying and releasing\n   dual-use models safely\, from how teams evaluat
 e a model's capabilities\n   before release to the release-strategy choice
 s that follow. Panelists\n   will weigh approaches and how those decisions
  shift as capabilities\n   advance. Looking ahead\, the conversation will 
 consider what the next\n   wave of increasingly capable models means for s
 afely deploying AI\n   models and how developers can prepare now for the e
 xponential to come.\n\n   Speakers:Emanuel Gawrieh\,Jason Clinton\,Bruce S
 chneier\,Heather Adkins\n\n   SpeakerBio:  Emanuel Gawrieh\, Co-Chair at A
 I Village\n\n   Emanuel Gawrieh is a Senior Chaos Containment Engineer on 
 the Advanced\n   Threat Protection team at Google\, where he specializes i
 n architecting\n   and proving out threat models for complex workloads in 
 highly\n   regulated environments. As a core member of Google's Secure AI\
 n   Framework (SAIF) team\, he works closely with Google AI Red Team\,\n  
  Google DeepMind\, and the Cloud CISO's office to establish AI security\n 
   standards - contributing to the early development and specialized\n   fi
 ne-tuning of dual-use models like SecGemini to mitigate emerging\n   AI-sp
 ecific risks. Outside of his work at Google\, he is a leading\n   voice in
  AI Security policy and community evangelism - having\n   presented at DEF
 CON\, RSA and is an active member of the [un]prompted\n   CFP board. Addit
 ionally\, he has architected the infrastructure for\n   several Generative
  AI Red Team events and actively advises\n   congressional leaders on AI a
 nd cybersecurity policy.\n\n   SpeakerBio:  Jason Clinton\, Deputy CISO at
  Anthropic\n\n   Jason joined Anthropic in April 2023 as its first CISO af
 ter more than\n   a decade at Google\, where he most recently led Chrome i
 nfrastructure\n   security working on defense against advanced persistent 
 threats. While\n   at Google\, he also worked on ChromeOS and Android Pay.
  At Anthropic\,\n   Jason guides security strategy including detection and
  response\,\n   compliance\, physical security\, security engineering and 
 IT. Jason\n   promotes the security organization's work to uphold Anthropi
 c's\n   Responsible Scaling Policy framework\, ensuring that the company h
 as\n   appropriate security safeguards in place for the responsible\n   de
 velopment and deployment of AI models. He is also the author of\n   "Ruby 
 Phrasebook".\n\n   SpeakerBio:  Bruce Schneier\, Advisory Board Member at\
 n   VerifiedVoting.org\n\n   Bruce Schneier is an internationally renowned
  security technologist\,\n   called a “security guru” by the Economist
 . He is the New York\n   Times best-selling author of 14 books – includi
 ng Rewiring Democracy\n   and A Hacker’s Mind -- as well as hundreds of 
 articles\, essays\, and\n   academic papers. His long-running newsletter a
 nd blog\, “Schneier on\n   Security\,” is one of the most popular sour
 ces of cybersecurity news\n   on the internet. Schneier is a Fellow and Le
 cturer in Public Policy at\n   the Harvard Kennedy School and the Munk Sch
 ool at the University of\n   Toronto. He is a fellow at the Berkman-Klein 
 Center for Internet and\n   Society at Harvard University\, a board member
  of the Electronic\n   Frontier Foundation and AccessNow\, and an advisory
  board member of\n   EPIC and VerifiedVoting.org. He is also the Chief of 
 Security\n   Architecture at Inrupt\, Inc.\n\n   SpeakerBio:  Heather Adki
 ns\, Vice President of Security Engineering at\n   Google\n\n   Heather Ad
 kins is a 24-year Google veteran and founding member of the\n   Google Sec
 urity Team. As VP\, Security Engineering and head of\n   Google’s Office
  of Cybersecurity Resilience she has built a global\n   team responsible f
 or maintaining the safety and security of Google’s\n   networks\, system
 s and applications. She has an extensive background in\n   practical secur
 ity\, and has worked to build and secure some of the\n   world’s largest
  infrastructure. She is co-author of Building Secure\n   and Reliable Syst
 ems (O’Reilly\, 2020)\, was deputy chair of CISA’s\n   Cyber Safety Re
 view Board\, and has advised numerous organizations on\n   how to adopt mo
 dern defendable architectures.\n\n   '\n\n   1. #LVCCW_Level1_Hall3\n\n\n
DTEND:20260807T215900Z
DTSTART:20260807T203000Z
LOCATION:DEF CON Talks - LVCCW Level 1 Hall 3 1006 (Main Track 1) and DCTV-
 1
SUMMARY:The Defender’s Dilemma: Releasing Dual-Use AI Models as Capabilit
 ies Climb
END:VEVENT
END:VCALENDAR