CyberSec&AI Connected

Ready when you are 😎

You miss us already? 😏 #CyberSecAI 2021 is over! Thanks for joining, we hope you enjoyed the conference. If you want more, check out our newest podcast. @CVUTPraha @Avast ava.st/3BV1Nw5

⏰The last workshop of this year's CyberSec&AI Connected is about to start! @mrigaki from @CVUTPraha is presenting: Stealing Malware Classification Models for Antivirus Evasion. #CyberSecAI

Up next, Zahra Ghodsi from @UCSanDiego will be talking about Safe and Private Learning: From Federated Edges to Fogs and Clouds. #CyberSecAI

Our next presenter is Professor Martine de Cock from the School of Engineering and Technology @UWTacoma. Hear her speak about Privacy-Preserving #MachineLearning with Secure Multiparty Computation. #CyberSecAI

Let's give a round of applause to today's moderator and a true expert in the #AI industry @fjbiondi! 👏 BTW, did you know Fabrizio is also a part of our Program Committee? Without him, #CyberSecAI wouldn't be what it is today. #ThankYou!

@maurapintor from @univca tells us how to understand why and how adversarial attacks fail. Do you have any guesses? #CyberSecAI

😈"Can we fool the classifier to think it's a goodware? Not so sure. Malware have a very complex structure and we might break the format if try." @zangobot from @univca reveals solutions to the problem in his research.

👨‍🍳Ugur Akyazi from @tudelft presents his "Hacking Menu - Ask your server about our specials". His work shows #CaaS makes cybercrime easily accessible to novice criminals with limited technical skills.

It's workshop day at #CyberSecAI! ⚡️ 11 experts will present their latest research today. Follow this thread all afternoon, and ask us your questions👇

🍕 "When i’m ordering food online, I want zero knowledge proof." - Darren Shou. What is it? 👇 Zero-Knowledge Proofs allow #data to be verified without revealing that data . It has the potential to revolutionize the way data is collected, used and transacted with. #CyberSecAI

#CTO Talks with 💥@Avast - Michal Pechoucek 💥@NortonLifelock - Darren Shou 💥@salesforce - Taher Elgamal 👇 #CyberSecAI 👇

Fabio Roli: "Adversarial examples remind us that #MachineLearning models are based on the independent and identically distributed assumption, which supposes that future data will resemble past data. This assumption creates problems in #cybersecurity." #CyberSecAI

Why is #MachineLearning so vulnerable? Fabio Roli explains: "#ML algorithms very often have large gradient learning function, which creates large sensitivity, meaning a small change of the input data can cause a big change of the predictions." #CyberSecAI

Up next on CyberSec&AI Connected, Fabio Roli from @univca is presenting live from Italy: #MachineLearning for #Cybersecurity: The Illusion of Progress? #CyberSecAI

💪 "It's still very much important to invest in people, training and security as well as #AI." - @jayabaloo

Our #security panel is on for the next 50 mins. Ask away! ⚡️ #CyberSecAI, #AI: what are the next revolutions, and where will they stop?

In today’s presentation at #CyberSecAI Connected, Nicholas Carlini tries to extract sensitive data out of a machine learning model. Is it possible? And what does it mean for the #privacy of our private #data? #CyberSecAI

In our “AI’s role in the future of personal data privacy” panel, we’re talking with experts about how we’ve progressed in #AI, #machinelearning, and #optimization. #CyberSecAI

#CyberSecAI Our very first speaker is @ssnstudy from @HeinzCollege, live from Pittsburgh - USA 🇺🇸. Want to learn about empirical research? Follow this thread!