Socket CISO Discusses Vendor Trust and AI Tooling Costs on Early Adopters Podcast

Socket Chief Information Security Officer Andrew Becherer recently appeared on the Early Adopters: CISO podcast, offering detailed insights into how security leaders evaluate new technology. Recorded at Black Hat 2026, the episode explores the challenges of securing organizations as artificial intelligence accelerates both software development and security operations.

Becherer emphasizes that while building a security program has become easier, securing an entire organization remains difficult. He draws on his experience at Datadog, where he started as the sole security employee in 2015. His tenure included responding to a 2016 breach, which fundamentally altered the company’s approach to security and informed his current strategies for program growth.

Context and Background

The discussion addresses how agentic coding influences the decision to build or buy security tools. AI can rapidly create incident-response tooling and analyze logs, but these prototypes introduce ongoing responsibilities. Teams must consider reliability, evolving requirements, and long-term maintenance, ensuring someone will support the software years after its initial creation.

Becherer notes that AI accelerates development but does not eliminate the need for judgment. Security leaders must assess whether a tool supports a stronger program or simply adds complexity. The conversation spans his unconventional path into cybersecurity and the challenges of securing AI agents, highlighting the importance of foundational work in supporting organizational growth.

Who Is Affected

Founders and security vendors are directly affected by Becherer’s advice on gaining buyer trust. He states that existing pain points and practical answers carry more weight than predictions about future problems. An honest account of security gaps builds confidence, whereas speculative warnings may undermine a deal. He shares a specific question he asks vendors to gauge their reliability and understanding of current security needs.

Security teams evaluating AI-built tools are also impacted. They must weigh the speed of AI-generated solutions against the long-term costs of maintenance. The conversation underscores the need for clear ownership and accountability, ensuring that AI-assisted development does not compromise the reliability of critical security infrastructure.

What should you do?

Security leaders should review their current tooling for long-term maintenance plans. Evaluate whether AI-generated tools have clear ownership and support structures. Founders should focus pitches on solving existing pain points rather than predicting future threats. Teams must assess the reliability of new technologies before integrating them into critical security workflows.

What to do and how to stay safe

• Audit existing security tools to ensure they have clear maintenance plans and long-term support.
• Evaluate AI-generated security solutions for reliability and accountability before deployment.
• Focus vendor pitches on solving immediate, documented security pain points rather than future risks.
• Implement a clear ownership model for all AI-assisted security tools to ensure ongoing updates.

Step-by-step guide: Prompt Injection Warning Signs: How to Spot AI Manipulation

General security guidance from the Threat Times newsroom. It is not confirmed advice from the organisations named in this story.

Frequently asked questions

What podcast did Socket CISO Andrew Becherer appear on?

He appeared on the Early Adopters