The Claude Code Security Summit

Опис

A six-hour virtual summit on 25 September 2026, starting at 1pm UTC, for teams securing Claude Code and other AI coding agents in production. The Claude Code Security Summit goes beyond prompting and product demos to examine the controls required when coding agents can inspect repositories, change multiple files, invoke external tools and remediate vulnerabilities. Organised by [Packt Publishing Limited](https://www.eventbrite.co.uk/o/70306584013), this vendor-neutral programme connects secure specification, adversarial testing, MCP trust boundaries and enterprise governance. ## What to expect Across four connected, practitioner-led sessions, attendees will follow the AI-assisted delivery lifecycle from initial requirements to production approval. The summit combines technical guidance with live demonstrations showing generated code, weaknesses, attack paths, remediation and the evidence needed at approval gates. Planned demonstrations and practical exercises include: - Building `SECURITY.md` and `CLAUDE.md` to constrain generated code - Connecting `REQUIREMENTS.md`, `ARCHITECTURE.md`, `SECURITY.md` and `CLAUDE.md` - Comparing the same feature with and without specifications - Turning feature requests into deterministic agent tasks with testable security criteria - Running model-led validation against explicit requirements - Attacking an AI-generated API or application, tracing the root cause and verifying the fix - Mapping identities, tools, data, permissions and third-party trust in an MCP workflow - Defining automated testing, security review and human sign-off gates - Drafting approval checklists covering acceptable autonomy and minimum production evidence ## Practitioner-led security expertise The speaker faculty brings together secure development standards, application security, ethical hacking and enterprise transformation:… [Register for this event](https://www.eventbrite.co.uk/e/the-claude-code-security-summit-tickets-1997159042458?aff=ebdssbdestsearch)