The prominent keyword at the RSAC™ 2025 Conference was community. This focus on community resonated throughout the week, echoed by speakers, attendees, volunteers, and in sessions.
Embracing Advanced Security Strategies
The central message was clear: cybersecurity is a collective endeavor. We cannot effectively combat increasingly sophisticated cybercriminals in isolation. In his presentation Canary in the Cloud Mine: Stealthy Tripwires to Detect Post-Breach Activity, Jenko Hwong, Threat Researcher at WideField Security noted, "Attacks are winning with stealthy techniques." It often feels like a relentless cycle: new, more sophisticated attacks emerge as quickly as we develop defenses. Even good security measures like strong passwords fortified with multi-factor authentication can be a challenge, as attackers are successfully bypassing these traditional security layers, so defenders need to think more proactively.
Highlighting the critical need for innovative security strategies, Andy Ozment, Chief Technology Risk Officer at Capital One, delivered his presentation, Dude, Where’s My Password? The Challenges of Getting to Passwordless. During his talk, Ozment asserted, "We are no fans of passwords; we've been struggling with them for decades, and now we have a compelling offer to move away from them," emphasizing the significance of adopting more robust, passwordless authentication techniques. He further detailed Capital One's successful deployment of device certification paired with FIDO2 passkeys to achieve passwordless authentication, noting the positive outcomes of decreased risk and enhanced user satisfaction.
Bridging Communication Gaps
James Tarala, Managing Partner at Cyverity, tackled a significant communication challenge in his presentation, Architecting Data Analytics for Continuous Risk Management. Tarala pointed out the differing needs of cybersecurity teams, who must communicate clearly with stakeholders, and leadership, who seek objective risk quantification. To reconcile these needs, Tarala presented the CRF Business Intelligence Model (BIM), a proposed architecture for sharing quantified risk data. The goal of this model is to equip stakeholders with the information needed to reduce uncertainty and make more informed risk mitigation choices.
In his presentation, Reporting Lines Matter: The 2025 CISO’s Place in the Org Chart, Todd Fitzgerald, CISO, Cybersecurity Leadership Author at CISO SPOTLIGHT LLC focused on six evolutionary stages and provided guidance on the security leader's reporting structure for 2025. A key message was, “Cybersecurity tone at the top can only be supportive if you are listening to the tone at the bottom,” emphasizing that organizational-wide communication and collaboration are crucial for strong cyber hygiene.
AI's Expanding Role in Cybersecurity
Yes, community was a central theme, but of course, AI was also prominent. In their presentation, The Always-On Purple Team: AI Agents on the Loose, Erik Van Buggenhout and Jeroen Vandeleur, both SANS Instructors, illustrated the evolution of AI in purple teaming. As their 2024 presentation concluded, LLMs could be a force multiplier for detection engineering, which proved to be true. Their 2025 presentation then focused on how agentic AI could be leveraged to fully automate purple team workflows. Each year, we can anticipate another layer of evolution in AI as more tools are invented.
The evolution of ChatGPT and its initial, less advanced capabilities compared to today's version was the focus of Mark Sherman's presentation, The Future of Secure Programming Using LLMs. Sherman, Director at CERT, Software Engineering Institute at CMU, addressed the considerable buzz around ChatGPT's potential to rewrite or generate new code for organizations, with some even forecasting a 70% reduction in their coding staff due to AI. However, Sherman cautioned that expert oversight of AI-generated code remains crucial to ensure its accuracy, as much of ChatGPT's output has demonstrated security vulnerabilities.
AI's double-edged nature for both attackers and defenders, coupled with a current lack of comprehensive understanding, was highlighted by Katerina Megas, Program Manager at NIST, in her presentation The Cybersecurity Framework and AI. Key concerns she raised included communication breakdowns between cybersecurity and AI teams, and the challenge for CISOs in managing novel AI risks alongside their existing responsibilities. Megas stressed the importance of shared responsibility, emphasizing that CISOs should foster collaboration across cybersecurity departments. This emphasis on organizational collaboration echoed in Kevin Mandia’s Keynote, Cybersecurity Year-in-Review and The Future Ahead. Mandia, along with co-presenter Nicole Perlroth, Cybersecurity Author, Investor & Advisor at Silver Buckshot Ventures, described effective leaders as consistently composed and motivating through dedication, underscoring the necessity of alignment and engagement from various departments in addressing security breaches.
Spotlighting Innovation and Research
As previously announced, the Top 10 RSAC™ Innovation Sandbox (ISB) Finalists received a $5 million uncapped SAFE investment. This year, the RSAC™ ISB content buzzed with words like "innovating," and each participant showcased their innovative tools.
RSAC™ 2025 Conference also expanded its offerings with the addition of tracks such as USENIX. This track brought together researchers, practitioners, system administrators, system programmers, and others interested in the latest advances in security.
Ultimately, RSAC™ 2025 Conference illuminated the path forward: a stronger, more resilient cybersecurity landscape forged through shared knowledge and collective action.