DSOLG May 2026 - AI incident response

## Details

Welcome to the DevSecOps London Gathering March Event on Wednesday 20 May. Along with our amazing talks we bring you the usual conversations, pizza and beer!

Location:

• 📍 Venue: Veracode, 36 Queen Street London, EC4R 1BN
• 📅 Date: Wednesday, 20 May 2026
• 🕕 Time: 6:00 PM – 8:00 PM

## Talk 1 - James Bevan

Abstract:
DevSecOps teams scan everything except the system governing their processes: Jira itself. When Jira configuration drifts, controls fail silently - workflows skip security gates, permissions grant unintended access, required fields disappear.
This session demonstrates how configuration drift creates security gaps and shows live detection across workflows, permissions, and automation. With agentic AI now modifying Jira configs, continuous monitoring of your control plane isn't optional.
Speaker Bio:
James is co-founder and CRO at Solcoro, building a platform that gives Enterprises and Atlassian Solution Partners real visibility into Jira Cloud environments. Alongside Solcoro, he also runs CROquet - a UK-based GTM consulting business. He takes on a small number of advisory and project engagements, focused on incentive design and revenue architecture. This combination lets him apply GTM theory in practice, both inside a SaaS business and across client environments.

## Talk 2 - Derrisa Tuscano

Abstract
The widespread deployment of Generative AI has introduced novel security incidents ranging from prompt injection and data exfiltration to misinformation cascades that traditional incident response frameworks are ill equipped to handle. This talk introduces a Practical Incident Response Framework for Generative AI Systems designed to bridge established cybersecurity standards with emerging AI assurance principles. I will discuss a response centric taxonomy that identifies six recurrent incident archetypes and presents a structured playbook aligned with NIST SP 800 61r3, NIST AI 600 1, MITRE ATLAS, and the OWASP LLM Top 10. Attendees will learn how to reduce the cognitive load on security teams during non deterministic AI failures by using tailored runbooks, RACI matrices, and Swimlane diagrams. Ultimately, the session will demonstrate how organizations can adapt traditional response models to operationalize AI incident response effectively.
Speaker Bio:
Derrisa Tuscano is a Cybersecurity postgraduate from the University of Warwick and a published researcher in Generative AI incident response. Her professional background includes leading numerous end to end project life cycles for Applied AI solutions at Quantiphi Analytics Solutions, where she specialized in architecting virtual agents within Google Cloud environments. Derrisa has been recognized with the WMG Excellence Scholarship and the Warwick Award for her commitment to skills development. Her work focuses on bridging the gap between enterprise cloud delivery and robust cyber risk management to ensure secure and compliant AI implementation.

##