AI Red Teaming LLM: Past, Present, and Future

Speaker Slides: https://static.sched.com/hosted_files...
Explore the world of AI Red Teaming Large Language Models (LLMs) - their origins, current challenges, and future possibilities. Since 2014, AI Red Teaming has been used to identify security risks in AI, mostly in computer vision. With advancements in ChatGPT and other LLMs, risks such as Prompt leakage, prompt injection, jailbreaks, poisoning, and logic manipulation attacks remain. As LLMs become more common in business applications, it is crucial to have AI Red Teaming skills which require expertise in computer hacking, ai, social engineering, psychology, neuroscience, mathematics, and logic.

Adarsh Nair
UST
Global Head - Information Security
India
https://adarshnair.com
Contact Me
  / adarshsvnair  
  / adarshsvnair  
  / toadarshnair  

Adarsh Nair serves as the Global Head - Information Security at UST. He is an information security strategist, author, keynote speaker and is a recognized Fellow of Information Privacy (FIP) of International Association of Privacy Professionals (IAPP). He has been recognized for his work in the Information Security sphere, including being awarded the Excellence Medal by the Hon’ble Chief Minister of Kerala for his outstanding contribution in the field of cybersecurity, as well as the top-contributor recognitions from Kerala State Police Chief, and Hall of Fame from Google. He has also authored a variety of articles, books and research papers to his credit.

Mohit Joshi
AppSentinels.ai
Co-Founder & CDO

Casey Ellis
Bugcrowd
Founder Chairman CTO

Casey Ellis, Founder, Chairman and CTO of Bugcrowd: Casey is a 20-year infosec veteran, servicing clients as a pen tester, security and risk consultant, solutions architect, and most recently as a career entrepreneur. Casey pioneered the Crowdsourced Security as a Service model, launching the first bug bounty programs on the Bugcrowd platform in 2012, and co-founded the disclose.io vulnerability disclosure standardization project in 2016.

Tom Brenan
CREST International
Executive Director
New Jersey
crest-approved.org
Contact Me
  / brennantom  
  / tombrennan  

Tom Brennan stands out as a distinguished cybersecurity executive and the U.S. division leader at CREST International, renowned for his profound impact on cybersecurity standards and practices. He excels in steering CREST's mission as a foremost accreditation body, ensuring robust cybersecurity protocols across the 16 critical infrastructure sectors as outlined by CISA.

In his pivotal role as Chairman of CREST USA, Tom spearheads major strategic initiatives. He is deeply committed to the promotion of accredited cybersecurity solutions and professional expertise, always with a keen focus on consumer privacy, security, and international safeguards. His proficiency spans a vast array of cybersecurity disciplines, including but not limited to, penetration testing, vulnerability assessment, application security, and threat intelligence. This breadth of knowledge empowers Tom to infuse CREST with insightful leadership and vision.

Tom's expertise extends to his role as Chief Information Officer at Mandelbaum Barrett, where he is entrusted with overseeing critical facets of infrastructure, privacy, and security operations. His counsel is sought after in numerous advisory capacities, including at the Gerson Lehrman Group and the New Jersey Institute of Technology. Additionally, his contributions as a member of the Cyber Fellows Advisory Council at NYU Tandon School of Engineering and the OWASP Foundation underscore his standing in the cybersecurity community.

Not only a driving force in his professional milieu, Tom is also a revered speaker, published author, and recognized expert in cybersecurity. His array of certifications, including CERT, CISSP, C|EH, and IAM, further endorse his vast expertise. Tom Brennan’s relentless dedication and pioneering spirit continue to propel the cybersecurity sector forward, firmly establishing him as a visionary and leader in safeguarding our digital world.

Managed by the OWASP® Foundation
https://owasp.org/