Hacking ML in images (and everywhere else)

This time we’re looking back into our archives to bring you a presentation from Guy Barnhart-Magen and Ezra Caltum. In their t2’18 talk the BSidesTLV co-founders cover offensive research possibilities when it comes to machine learning systems. Do you know which ML attacks have the most business impact? Watch the video to learn more.

The presentation will be answering questions such as “what does it mean to hack a machine learning system?” and “what would you actually target?”, with an emphasis on the methodology and the way Guy and Ezra approached the problem.

We have always enjoyed these types of talks, as the shared knowledge powers the audience to do research and find their own zero days. Speaking of research and zero days – don’t forget to checkout BSidesTLV, coming July 2 2020!

Physical Penetration Testing

An audience favorite from the archives – Walter Belgers with “Physical Penetration Testing”. This a great example of a presentation, which hasn’t aged a day since as updates to physical security controls rarely happen at regular predefined intervals. In addition to being good fun, covert entry can also save time and effort if you happen to *krhm* forget your keys.

The content was tailored to include material specific to Finland.  Impatient viewers just wanting to see the Abloy Classic lockpicking can start the video from 47:07. Those watching the presentation more carefully might notice the mention of a familiar name who presented at t2’17.

Cosmic Duke

In this video from t2’15 Artturi Lehtiö peers inside over 7 years of state sponsored malware operations. The presentation covers themes such as less glamorous side of APT research, tools and approaches, in addition to the challenges related to publishing this type of information.

Those into offensive work can view the video as training material for improving tradecraft.

For additional information on the subject, here are the links for the whitepapers mentioned during the talk.

Surviving the times

The beauty of good research is that it rarely gets old over night. Instead it matures, giving perspective to fundamentals or an outlook to a different time – possibly even igniting or re-igniting the passion and serving as a platform for fresh ideas. New technologies fall in the face of old methods yet defensive tactics still work despite being exposed to the real world for decades/centuries/millennia.

In 2015 we recorded talks for the first time, and continued doing so last year. These will be released semi-regularly, in no particular order, in the coming months. Having had the privilege of witnessing world class content at t2 over the years, we know what stands the test of time. While memes and movie references can either become classics or disappear completely, talent and fruits of persistent work remain.

Conferences are all about meeting people, making connections and exchanging information. By showcasing past presentations we hope to inspire you to start doing research and tinkering with technology – you will meet like-minded people from all over the world at t2.