Webinar Description
Key Takeaways
- Explores how autonomous AI agents are fundamentally changing software development workflows and creating new security challenges
- Introduces Agentic Development Security (ADS) as a framework for securing AI-driven development environments
- Addresses the limitations of traditional application security models when applied to agent-driven workflows
- Designed for AppSec leaders, security architects, platform engineers and DevSecOps professionals
- Covers practical approaches to maintaining visibility, governance and control over AI-generated code
Introduction
As AI agents transition from passive coding assistants to autonomous systems capable of executing workflows, invoking tools and generating production-ready code, the security landscape for software development is undergoing a fundamental shift. This live webinar from Snyk examines the emerging discipline of Agentic Development Security and its role in helping organisations navigate the complexities of securing AI-driven development at scale. The session is aimed at application security professionals, platform engineers and technical leaders grappling with how to maintain robust security postures without constraining the velocity that AI-powered development promises.
About This Event
Titled “Securing AI-Driven Software Development in the Age of AI Agents,” this virtual webinar brings together product and engineering leaders from Snyk to address one of the most pressing challenges facing modern development organisations. The session moves beyond theoretical discussion to offer practical guidance on implementing security controls that accommodate the unique characteristics of agentic development environments.
The webinar format allows for focused exploration of how traditional application security assumptions break down when software is increasingly authored by autonomous systems operating at machine speed rather than human pace.
Why Traditional Application Security Models Are Struggling
Conventional application security frameworks were designed around human-centric development workflows. Code reviews, pull request approvals and periodic security scans assume that humans remain the primary authors and decision-makers throughout the software development lifecycle. AI agents challenge these assumptions in several fundamental ways.
When an AI agent autonomously generates code, interacts with external systems and executes multi-step workflows, the traditional checkpoints that security teams rely upon become insufficient. The volume and velocity of AI-generated code can overwhelm manual review processes, while the opacity of agent decision-making complicates efforts to understand why particular code patterns or dependencies were chosen.
This webinar examines these breakdowns in detail, helping attendees understand where existing security investments remain valuable and where new approaches are required.
Understanding Agentic Development Security
Agentic Development Security represents an emerging framework for addressing the unique risks introduced when AI agents become active participants in software creation. Rather than treating AI-generated code as simply another input to existing security pipelines, ADS recognises that the agents themselves—their tools, behaviours and supply chains—require dedicated security consideration.
The webinar explores how ADS extends security controls into AI development workflows without creating friction that negates the productivity benefits organisations seek from AI adoption. This balance between security rigour and development velocity sits at the heart of the discussion.
Central to the ADS approach is the concept of Coding Agent Security, which focuses on real-time governance of agent behaviours and securing the agent supply chain. Just as organisations learned to scrutinise open source dependencies for vulnerabilities, they must now consider the provenance and trustworthiness of the AI agents and tools integrated into their development environments.
Risks in Agent-Driven Development Environments
The session addresses several categories of risk that emerge when AI agents take on expanded roles in software development. These include risks from the tools that agents invoke, the behaviours agents exhibit during autonomous operation, and the code and artefacts that agents produce.
AI-generated code may introduce vulnerabilities that differ in character from those typically created by human developers. Agents may select dependencies based on training data that predates the discovery of critical vulnerabilities, or they may generate code patterns that technically function but violate security policies or compliance requirements. Without appropriate visibility into agent activities, security teams may struggle to detect these issues until they manifest in production environments.
The webinar also considers supply chain risks specific to agentic development. The plugins, extensions and tool integrations that expand agent capabilities represent potential attack surfaces that traditional application security scanning may not adequately address.
Maintaining Visibility, Governance and Control
A recurring theme throughout the webinar is the challenge of maintaining meaningful oversight in environments where software creation increasingly occurs at machine speed. The session explores practical approaches to achieving visibility into what AI agents are doing, establishing governance frameworks that guide agent behaviour, and implementing controls that can intervene when necessary without creating bottlenecks.
Snyk Studio features in this discussion as a mechanism for ensuring that AI-generated code meets organisational trust and security requirements. The platform aims to provide the instrumentation necessary for security teams to understand and manage AI-driven development activities.
Effective governance in agentic environments requires rethinking where and how security decisions are made. Rather than relying solely on post-hoc scanning and review, organisations may need to embed security considerations into the agent workflows themselves, creating guardrails that operate in real time.
Who Should Attend
This webinar is designed for technical leaders responsible for securing software development environments, particularly those whose organisations are adopting or scaling AI-driven development practices. The content is most relevant for:
- Chief Information Security Officers evaluating the security implications of AI agent adoption
- Security architects designing controls for AI-augmented development pipelines
- Application security managers adapting programmes to address AI-generated code
- DevSecOps professionals integrating security into increasingly automated workflows
- Platform engineers building and maintaining AI-enabled development environments
Attendees will benefit most if they have existing familiarity with application security concepts and are actively considering or implementing AI coding assistants and agents within their organisations.
The Broader Industry Context
The questions this webinar addresses reflect broader industry uncertainty about how to govern AI systems that operate with increasing autonomy. As AI agents become more capable—moving from suggesting code snippets to orchestrating complex development tasks—the security community is working to establish frameworks, best practices and tooling appropriate to this new reality.
Organisations that delay consideration of these issues risk accumulating technical and security debt as AI-generated code proliferates through their systems. Conversely, those that implement overly restrictive controls may find themselves unable to realise the productivity benefits that drive AI adoption in the first place. Finding the appropriate balance requires understanding both the risks and the practical mechanisms available for managing them.