Glasswall at USENIX Security 2024

Recently, Dr. Aqib Rashid, our Senior Data Scientist, was in Philadelphia, USA, and attended USENIX Security 2024, a top-tier cybersecurity conference. It offered a great opportunity to engage with experts in the field of cybersecurity and machine learning research, introduce and promote Glasswall CDR and build some great connections. Here, he shares his perspectives on this important event.

In cybersecurity and machine learning research, where Glasswall is focusing significant time and investment, there are various conferences considered the most prestigious and top-tier; USENIX Security is one of them. It’s been going strong for over 20 years and has showcased state-of-the-art work in cybersecurity and, more recently, security in Machine Learning (ML). Having a research paper accepted at this venue is regarded as a significant achievement in the field. The conference brings together researchers and practitioners from both industry and academia to showcase the latest state-of-the-art developments in cybersecurity and machine learning.

This year, in the opening session, there was a talk by David Brumley, who is a former professor but now runs a company called ForAllSecure, Inc., which produces a product known as Mayhem. It’s an AI-driven AppSec platform, which is used by the Department of Defense. The talk was especially interesting as Brumley showcased several key lessons he learnt when transitioning from academia to industry, such as “don’t ask the user to do work before showing value” and “avoid technically correct but useless answers”. He also introduced a capture-the-flag platform, picoCTF, which appeals to all types of audiences seeking to enhance their cybersecurity knowledge.

Technical sessions

The conference comprised multiple tracks, which can be considered high-level topics, such as "Security for ML Systems" or “Network Intrusion Detection Systems”, which are then split into technical sessions. Novel work and research papers are presented in different technical sessions within these tracks.

As described, the technical sessions are the core part of the conference, where researchers and industry practitioners present their novel work. All of the technical sessions are available here, and although it is difficult to summarize all of the content, a huge emphasis was placed on Large Language Models (LLMs) and their potential for use in security domains and applications. This is also an area where the Advanced Research Team at Glasswall is focusing its efforts, developing concepts and ideas for future work.

Throughout the event, I also devoted a significant amount of time to networking and introducing partners and experts to CDR, its capabilities and what Glasswall offers. Many were impressed by not only the technology, but our client base, our state-of-the-art research and how well-established our technology is.

Advanced research and innovation

In addition, many researchers have expressed interest in experimenting with CDR, and as part of our ongoing engagement with the cybersecurity community, Glasswall has been actively collaborating with academia with visits to various institutions and organizations. We also have some significant interest from top-tier institutions in working with us on studying CDR further.

This forms part of a strategic commitment to the positive impact research and innovation into malicious documents can have on our customers and the wider cybersecurity community. The growing Glasswall Advanced Research Team brings together recognized industry experts and pioneers in the development of groundbreaking CDR solutions with a strong focus on the integration of AI and ML technologies.

Overall, USENIX Security 2024 was an extremely fruitful and beneficial conference for both myself and Glasswall, and I look forward to attending more events such as this in the future.

‍