AI, cybersecurity, and quantum innovations at DBSec'24

In today's digital world, nothing beats the spark of in-person collaboration. The 38th annual International Federation for Information Processing Working Group (IFIP WG) 11.3 Conference on Data and Applications Security and Privacy (DBSec'24) proved this point brilliantly. From July 15-17, 2024, Cisco's headquarters in San Jose became the epicenter of artificial intelligence (AI) and cybersecurity innovation as brilliant minds from across the globe gathered to shape the future of our digital defenses.

A global brain trust 

Imagine a room buzzing with the world's top AI, quantum, and cybersecurity experts — that was DBSec'24. Hosted by Cisco Research and co-chaired by Outshift by Cisco's  Ashish Kundu along with Haibing Lu from Santa Clara University, Ram Krishnan from the University of Texas at San Antonio, and Anna Lisa Ferrara from the University of Molise, Italy, the conference brought together thought leaders from prestigious institutions worldwide. Attendees represented a truly global perspective, hailing from Italy, France, Singapore, India, the United States, and Kuwait. This diverse mix of academia, industry, and government created a unique melting pot of ideas, enriching discussions and fostering a truly international exchange of ideas.

Outshift's comprehensive approach to AI and cybersecurity 

Outshift's presence was felt strongly through three insightful sessions that showcased its holistic approach to AI and cybersecurity: 

Navigating the AI frontier 

Vijoy Pandey, Ph.D., Senior Vice President of Outshift, delivered a keynote that delved into Outshift's incubation and innovation journey. He shared Outshift's process for identifying promising ideas, rapidly building and validating minimal viable products, and strategically launching and scaling successful solutions. Pandey highlighted Panoptica and Motific — two products the team has successfully launched — while also touching on some incubation ideas that didn't succeed, emphasizing the valuable lessons learned from these experiences.

Pandey emphasized Outshift's focus on the interplay of automation and abstraction as key axes of innovation, illustrating how advancing along these dimensions empowers new ways of thinking and problem-solving. Looking to the future, he introduced Outshift's vision for an "agentically driven framework" that could fundamentally change how work is approached across industries. 

During the Q&A, Pandey addressed the balance between top-down and bottom-up innovation. He advocated for a hybrid approach: defining focus areas from the top while encouraging bottom-up ideation within these lanes. Open source projects, he suggested, offer an ideal middle ground for academia-industry collaboration, bridging theoretical research with practical implementation. 

The future of AI, GenAI, and cybersecurity innovation

Papi Menon, Vice President of Product at Outshift, participated in a panel discussion that explored the complex interplay between AI advancements and cybersecurity challenges. He highlighted key points on the dual nature of AI in cybersecurity, both enhancing security measures and creating new vulnerabilities. Menon also shared the challenges in enterprise AI adoption, including issues of trust and explainability and the evolving landscape of cyber threats, particularly in areas like social engineering. He concluded that there is a critical need for ongoing research, innovation, and cross-sector collaboration to address these emerging challenges. 

Cybersecurity in the age of AI and quantum 

Ramana Kompella, Ph.D., Head of Cisco Research, set the stage with his talk on the profound impact of AI and quantum technologies on cybersecurity. He highlighted the rapid maturation of these fields and the unprepared state of current cybersecurity measures. Kompella detailed Cisco Research's efforts in addressing key challenges and emphasized the need for innovative solutions that can keep pace with emerging threats in this new technological landscape. 

These sessions collectively demonstrated Outshift's commitment to driving innovation while addressing the complex challenges at the intersection of AI and cybersecurity. 

Emerging themes and groundbreaking research 

DBSec'24 showcased cutting-edge research at the intersection of AI and cybersecurity. Some of the standout emerging themes and groundbreaking research were around the relationship between data and AI trustworthiness; data ownership, licensing, and security in AI pipelines; and novel attack vectors on large language models (LLMs). 

Three notable presentations in this area were:  

• "ALERT: A Framework for Efficient Extraction of Attack Techniques from Cyber Threat Intelligence Reports Using Active Learning," by researchers from NIST and UT Dallas, showcasing the power of collaborative research. 
• "All Your LLMs Belong To Us: Experiments with a New Extortion Phishing Dataset," by Fatima Zahra Qachfar, demonstrating innovative attacks on LLMs. 
• "A Graph-based Framework for ABAC Policy Enforcement and Analysis," by Mian Yang, which earned the Best Student Paper award. 

DBSec'24 also honored outstanding contributions to the field: 

• Outstanding Research Award – R. K. Shyamasundar, IIT Bombay, India 
• Outstanding Contribution Award - Cliff Wang, National Science Foundation, U.S. 
• Best Paper Award - "DT-Anon: Decision Tree Target-Driven Anonymization" by Sabrina De Capitani di Vimercati, Sara Foresti, Valerio Ghirimoldi and Pierangela Samarati 
• Best Student Paper Award - "A Graph-based Framework for ABAC Policy Enforcement and Analysis" by Mian Yang, Vijayalakshmi Atluri, Shamik Sural and Jaideep Vaidya

Forging the future, together: A partnership between industry and academia 

DBSec'24 wasn't just about presentations — it was about connections. Spontaneous discussions in hallways and over coffee breaks led to new collaborations that could shape AI and cybersecurity for years to come. The conference also highlighted the strength of industry-academia partnerships. Several presenting professors have been supported by research gifts from Cisco Research, demonstrating the real-world impact of these collaborations. 

As we face ever-evolving cyber threats, gatherings like DBSec'24 remind us of the power of human collaboration. By bringing together diverse perspectives and fostering open dialogue, we're not just reacting to challenges — we're actively shaping a more secure digital future for all. 

Looking ahead

Reflecting on DBSec'24, it's clear that the future of data security and AI will be shaped by the seeds planted at this conference. Topics likely to dominate future discussions include ensuring trustworthy data for AI applications, the potential of synthetic data in AI development, protecting LLMs from novel attack vectors, and leveraging graphs for enhanced security measures. 

As we look ahead to next year's DBSec in Europe, we carry forward the insights, connections, and inspirations gained from this remarkable gathering in Silicon Valley, along with the recognition of these exceptional contributions to the field of data and applications security and privacy. 

Did you miss DBSec’24? Connect with Outshift at upcoming industry events and learn more about how Cisco Research is collaborating with world-class academic researchers to investigate new technologies.