Join our Mailing list!
Get all the latest news, exclusive deals, and feature updates.
CSA · SANS · OWASP GenAI Security Project · London Summit
SDLC Security Guardrails Summit — London
250 CISOs wrote the "AI Storm" strategy paper — now CSA is bringing the room back together to present the follow-up. Third summit in the first wave, after San Francisco, New York and DC.
Phoenix Security is a sponsor & content partner of this summit — organized by CSA, hosted by Telefónica Tech.
On the day
One afternoon, one room, no slide decks selling a product. Practitioners on threat-led triage, engineering the agentic SDLC, and where the vulnerability landscape goes next.
Framing for the afternoon: why this follow-up series exists, and what the room is here to work on.
In March 2026 the Linux kernel’s AI-generated bug reports went from 2 to 10 a week and stopped being hallucinated. Every one was valid. curl, which had killed its bug bounty over AI slop, was now getting genuine AI-assisted findings. Discovery had flipped from a noise problem to a volume problem.
That makes it a triage problem. The question is no longer whether a bug is real. It is which of the hundreds to fix first, and a severity score cannot answer that. A score rates a flaw in isolation. It doesn’t know whether an attacker can reach that code in your architecture, what boundary the path crosses, or what it chains into. Those are threat-model questions, not severity ones.
This talk is that method, proven on real code. Point frontier AI at a real application and it surfaces dozens of true vulnerabilities with no order. Rank the same set by reachability and chaining against a live threat model of the system and the order changes: what an attacker can reach without a session moves up, an apparently severe finding that never crosses a trust boundary drops down the list, and two or three minor flaws no one singled out turn out to chain into a serious exploit no per-finding score could see. Matt runs it with the tooling he built for exactly this: STRIDE-GPT CLI generates the threat model straight from the codebase, and a triage skill applies the ranking with the reasoning attached. The method carries to any model or finding source, but STRIDE-GPT is the reference implementation, and is open source. This talk is a practical companion to the AI Vulnerability Storm briefing, and one worked answer to how you triage when the volume arrives.
We took our own engineering from 40 commits a week to 800 with the same headcount. The interesting problem was never how to go faster. It was how to hold that rate without shipping the 45% of AI-generated code that carries an OWASP Top-10 flaw — a figure that has stayed flat across every model generation since 2023. This brief is what we changed, and why the controls that kept us secure turned out to be the same controls that cut the token bill rather than adding to it.
We wanted to build context and security in one place and avoid token burning, we worked with knowledge graph and localized context, auto healing documentation and auto lesson learned. We ended up saving an average of 10 to 33X the token in development as well as the issue captured at PR level.
The argument is structural. Security belongs in the repo, not bolted on after the PR. Four versioned rule files give the agent an authoritative folder map, a known stack, and the security constraints for the code it is touching. One line of repo-level rule changes the default output from a raw injection sink to a sanitised one. Scanning moves to source, where the fix is still free, and the PR gate runs a bounded six-point check on a cheap model rather than an open-ended review on an expensive one. Underneath sits the ordering that makes any of this affordable: deterministic scanners first, graph context second, reasoning last — which is a security decision and a cost decision at the same time.
Key discussion points
Explore the rush to adopt AI, and what gets missed when speed outruns security, governance and good design. Move from ad-hoc usage, to experimentation and scalable design. You’ll leave with actionable design patterns to use, or adapt for your own environment.
Francesco Cipollone, Sean Turner & panelists · more TBC
Supply chain risk, AI applied to everything from discovery to remediation, and the open question of who owns the AI security budget.
All speakers & attendees
Drinks and conversation to close out the afternoon.
Agenda and speaker line-up as shared by the organizers ahead of the event — subject to change.
Approval required · limited seats
Who’s behind this summit
The summit is organized by the Cloud Security Alliance (CSA), together with SANS and the OWASP GenAI Security Project, as a direct follow-up to the "AI Storm" strategy paper led by Gadi Evron, Rich Mogull and Rob T. Lee. Telefónica Tech is hosting the London session at its offices, and Phoenix Security is sponsoring the summit and speaking on securing the agentic SDLC.
Phoenix Security is an Actionable ASPM platform. It turns fragmented vulnerability data from code, containers and cloud into one prioritized, reachability-aware remediation backlog — instead of another dashboard full of unranked findings. Contextual deduplication and a 4D risk model are built for exactly the problem this summit is about: making AI-scale discovery something a security team can actually act on.
phoenix.securityTelefónica Tech is Telefónica’s global technology business, delivering cybersecurity, cloud, IoT, big data and AI services to enterprise customers in more than 170 countries. Its UK & Ireland cybersecurity and cloud team is hosting the London summit at its offices in the City of London.
telefonicatech.com
Derek Fisher – Head of product security at a global fintech. Speaker, instructor, and author in application security.
Derek is an award winning author of a children’s book series in cybersecurity as well as the author of “The Application Security Handbook.” He is a university instructor at Temple University where he teaches software development security to undergraduate and graduate students. He is a speaker on topics in the cybersecurity space and has led teams, large and small, at organizations in the healthcare and financial industries. He has built and matured information security teams as well as implemented organizational information security strategies to reduce the organizations risk.
Derek got his start in the hardware engineering space where he learned about designing circuits and building assemblies for commercial and military applications. He later pursued a computer science degree in order to advance a career in software development. This is where Derek was introduced to cybersecurity and soon caught the bug. He found a mentor to help him grow in cybersecurity and then pursued a graduate degree in the subject.
Since then Derek has worked in the product security space as an architect and leader. He has led teams to deliver more secure software in organizations from multiple industries. His focus has been to raise the security awareness of the engineering organization while maintaining a practice of secure code development, delivery, and operations.
In his role, Jeevan handles a range of tasks, from architecting security solutions to collaborating with Engineering Leadership to address security vulnerabilities at scale and embed security into the fabric of the organization.
Jeevan Singh is the Director of Security Engineering at Rippling, with a background spanning various Engineering and Security leadership roles over the course of his career. He’s dedicated to the integration of security practices into software development, working to create a security-aware culture within organizations and imparting security best practices to the team.
In his role, Jeevan handles a range of tasks, from architecting security solutions to collaborating with Engineering Leadership to address security vulnerabilities at scale and embed security into the fabric of the organization.
James Berthoty has over ten years of experience across product and security domains. He founded Latio Tech to help companies find the right security tools for their needs without vendor bias.
Chris Romeo is a leading voice and thinker in application security, threat modeling, and security champions and the CEO of Devici and General Partner at Kerr Ventures. Chris hosts the award-winning “Application Security Podcast,” “The Security Table,” and “The Threat Modeling Podcast” and is a highly rated industry speaker and trainer, featured at the RSA Conference, the AppSec Village @ DefCon, OWASP Global AppSec, ISC2 Security Congress, InfoSec World and All Day DevOps. Chris founded Security Journey, a security education company, leading to an exit in 2022. Chris was the Chief Security Advocate at Cisco, spreading security knowledge through education and champion programs. Chris has twenty-six years of security experience, holding positions across the gamut, including application security, security engineering, incident response, and various Executive roles. Chris holds the CISSP and CSSLP certifications.
Jim Manico is the founder of Manicode Security, where he trains software developers on secure coding and security engineering. Jim is also the founder of Brakeman Security, Inc. and an investor/advisor for Signal Sciences. He is the author of Iron-Clad Java: Building Secure Web Applications (McGraw-Hill), a frequent speaker on secure software practices, and a member of the JavaOne Rockstar speaker community. Jim is also a volunteer for and former board member of the OWASP foundation.
Get all the latest news, exclusive deals, and feature updates.
