Chainguard webinar addresses security at AI-driven development speed

Chainguard has released a webinar, The True Cost of Building at Machine Speed, focused on securing AI-assisted software development when teams can produce 10 to 50 times more code. The session examines how security organisations can maintain control over vulnerabilities, dependencies and production risk as development output accelerates.

The central issue is not simply whether individual pieces of AI-generated code are secure. It is whether established application-security processes can cope when the volume of components, findings and required fixes grows more quickly than people can review and remediate them.

More output can create a larger security backlog

For years, many teams have followed a familiar cycle: developers write code, scanners identify issues, security staff prioritise them, and engineers address the most important findings. AI puts pressure on that workflow because faster output can also mean more dependencies, more software components and more vulnerability records to manage.

Chainguard argues that adding scanning capacity alone does not resolve the problem. Without a way to make decisions and apply controls at the pace software is built, additional findings may simply create a larger backlog. Meanwhile, the AI models that help developers create and understand software are also available to attackers, increasing the pressure on security teams.

Controls must work before production

The webinar explores where CVE-driven remediation begins to struggle at machine scale, what secure-by-default development should involve, and how organisations can establish guardrails before code reaches production. It also considers the expanding software attack surface and the limits of existing vulnerability-management processes as AI adoption grows.

The wider context includes AI agent incidents and vulnerability chains, where incidents involving AI agents sit alongside vulnerabilities and social-engineering delivery chains, illustrating why software security cannot be treated as an isolated engineering concern.

Governance becomes a leadership question

AI-assisted development is increasingly a governance issue as well as an engineering choice. Security leaders need clarity on who owns the risk, what degree of exposure the organisation accepts, and how those decisions can be explained to executives and boards.

The practical implication for businesses is to avoid treating slower development as the primary safeguard. Instead, they need secure-by-default controls and accountable risk management designed for current development practices, so security can scale alongside AI-enabled delivery.