BEGIN:VCALENDAR VERSION:2.0 PRODID:Data::ICal 0.24 BEGIN:VEVENT DESCRIPTION: 'Title: Growing the Community of AI Hackers with the Generat ive Red\n Team\n When: Friday\, Aug 11\, 09:00 - 09:45 PDT\n Where: Caesars Forum - Academy - 407-410 - Track 4 - [1]Map\n Speakers:Austin C arson\,Rumman Chowdhury\,Sven Cattell\n\n SpeakerBio:Austin Carson \, Fo under & President at SeedAI\n Austin Carson is the Founder and President of SeedAI\, a nonprofit\n established to work with a diverse group of p olicymakers\, academics\,\n and private sector experts to help communiti es across the United\n States access the resources they need to engage w ith AI.\n\n Previously\, Austin established and led the DC government af fairs\n operation for NVIDIA\, translating NVIDIA’s expertise in artif icial\n intelligence and high performance computing for policymakers. Pr ior to\n joining NVIDIA\, Austin held a number of public sector and NGO\ n positions\, including serving as Legislative Director for Chairman\n Michael McCaul and Executive Director for the Technology Freedom\n Inst itute. Austin co-founded the Congressional Tech Staff Association\,\n co -led the Congressional Cybersecurity Caucus and the Congressional\n High Tech Caucus\, and is a founding fellow of the Internet Law and\n Policy Foundry.\n\n\n SpeakerBio:Rumman Chowdhury \, Co-Founder and CEO at Hum ane\n Intelligence\n Dr. Rumman Chowdhury’s passion lies at the inte rsection of\n artificial intelligence and humanity. She is a pioneer in the field of\n applied algorithmic ethics\, creating cutting-edge socio- technical\n solutions for ethical\, explainable and transparent AI. Dr. Chowdhury\n currently runs the nonprofit Humane Intelligence and is a Re sponsible\n AI Fellow at the Berkman Klein Center for Internet & Society at\n Harvard University.\n\n Previously\, Dr. Chowdhury was the Direc tor of META (ML Ethics\,\n Transparency\, and Accountability) team at Tw itter\, leading a team of\n applied researchers and engineers to identif y and mitigate algorithmic\n harms on the platform. Prior to Twitter\, s he was CEO and founder of\n Parity\, an enterprise algorithmic audit pla tform company. She formerly\n served as Global Lead for Responsible AI a t Accenture Applied\n Intelligence. In her work as Accenture’s Respons ible AI lead\, she\n led the design of the Fairness Tool\, a first-in-in dustry algorithmic\n tool to identify and mitigate bias in AI systems.\n \n Dr. Chowdhury is dedicated to cultivating and growing the next wave o f\n technology-forward companies enabling the responsible use of emergin g\n technologies.\n\n\n SpeakerBio:Sven Cattell \, Founder at nbhd.ai & AI Village\n Sven founded the AI Village in 2018 and has been running it ever\n since. Sven is also the founder of nbhd.ai\, a startup focused on the\n security and integrity of datasets and the AI they build. He w as\n previously a senior data scientist at Elastic where he built the\n malware model training pipeline. He has a PhD in Algebraic Topology\,\n and a postdoc in geometric machine learning where he focused on\n anom aly and novelty detection.\n Twitter: [2]@@comathematician\n\n Descrip tion:\n We’re running the largest live AI hacking event ever in the AI \n village this year. Anthropic\, Google\, HuggingFace\, Meta\, NVIDIA\, \n OpenAI\, and Stability\, have all provided models to attack and Scale AI\n have built the platform. This event is orders of magnitude bigger than\n any previous AI red team effort. There are observers from the Whi te\n House\, NIST\, NSF\, and the EU coming to learn from hackers. We bu ilt\n this event to grow the community that knows how to effectively\n evaluate Large Language Models as it is much more than prompt\n injecti ons and jailbreaks.\n\n AI works fundamentally differently to traditiona l software and only\n forms a part of a product. Trust and Security of A I in a system thus\n has to work fundamentally differently to traditiona l software. This is\n especially true for generative AI systems. The cor e difference is AI\n is a stochastic component of software and is allowe d to make a small\n amount of mistakes. This changes bug hunting\, repor ting\, and payouts.\n\n Come to this talk to hear about how and why we o rganized this\, and the\n history of algorithmic & bias bounties that le d up to the largest one\n ever at DEFCON 31. We’ll also give you some tips to help you in the\n contest.\n\n REFERENCES\n\n We Need Bug Bo unties for Bad Algorithms - Amit Elazari - [3]https://www.vice.com/en/arti cle/8xkyj3/we-need-bug-bounties-for-bad-algorithms\n\n Introducing Twitt er’s first algorithmic bias bounty challenge -\n Rumman Chowdhury & Ju tta Williams - [4]https://blog.twitter.com/engineering/en_us/topics/insigh ts/2021/algorithmic-bias-bounty-challenge\n\n Sharing learnings from the first algorithmic bias bounty challenge -\n Kyra Yee & Irene Font Perad ejordi - [5]https://blog.twitter.com/engineering/en_us/topics/insights/202 1/learnings-from-the-first-algorithmic-bias-bounty-challenge\n\n Bias Bu ccaneers - Rumman Chowdhury\, Jutta Williams\, Subho Majumdar\,\n Scott Steinhardt\, Ben Colman - [6]https://www.biasbuccaneers.org/\n\n An Algo rithmic Framework for Bias Bounties - Ira Globus-Harris\,\n Michael Kear ns\, Aaron Roth - [7]https://arxiv.org/abs/2201.10408\n\n Machine Learni ng Security Evasion Competition - Hyrum Anderson\, et al.\n - [8]https:/ /mlsec.io/ [9]https://cujo.com/announcing-the-winners-of-the-2021-machine- learning-security-evasion-competition/\n\n MITRE ATLAS - Ram Shankar Siv a Kumar\, et al - [10]https://atlas.mitre.org/\n\n The Spherical Cow of ML Security - Sven Cattell - [11]http://aivillage.org/adversarial%20ml/sph erical-cow/\n\n The Case for a Hippocratic Oath for Connected Medical De vices:\n Viewpoint - Beau Woods\, Andrea Coravos\, and Joshua David Corm an - [12]https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6444210/\n\n Annou ncing OpenAI’s Bug Bounty Program - [13]https://openai.com/blog/bug-boun ty-program\n\n Microsoft Malware Classification Challenge - Royi Ronen\, Marian Radu\,\n Corina Feuerstein\, Elad Yom-Tov\, Mansour Ahmadi - [14 ]https://www.kaggle.com/c/malware-classification\n [15]https://arxiv.org /abs/1802.10135\n\n '\n\n 1. #CaesarsAcademyBR\n 2. https://twitter. com/@comathematician\n 3. https://www.vice.com/en/article/8xkyj3/we-need -bug-bounties-for-bad-algorithms\n 4. https://blog.twitter.com/engineeri ng/en_us/topics/insights/2021/algorithmic-bias-bounty-challenge\n 5. htt ps://blog.twitter.com/engineering/en_us/topics/insights/2021/learnings-fro m-the-first-algorithmic-bias-bounty-challenge\n 6. https://www.biasbucca neers.org/\n 7. https://arxiv.org/abs/2201.10408\n 8. https://mlsec.io /\n 9. https://cujo.com/announcing-the-winners-of-the-2021-machine-learn ing-security-evasion-competition/\n 10. https://atlas.mitre.org/\n 11. http://aivillage.org/adversarial%20ml/spherical-cow/\n 12. https://www. ncbi.nlm.nih.gov/pmc/articles/PMC6444210/\n 13. https://openai.com/blog/ bug-bounty-program\n 14. https://www.kaggle.com/c/malware-classification \n 15. https://arxiv.org/abs/1802.10135\n\n\n DTEND:20230811T164500Z DTSTART:20230811T160000Z LOCATION:DC - Caesars Forum - Academy - 407-410 - Track 4 SUMMARY:Growing the Community of AI Hackers with the Generative Red Team END:VEVENT END:VCALENDAR