Datadog's Former CISO: AI Security Tools' Hidden Maintenance Costs Demand Vendor Transparency

Socket CISO Andrew Becherer recently appeared on the Early Adopters: CISO podcast to discuss the evolving challenges of cybersecurity leadership, vendor trust, and the integration of artificial intelligence into security operations. Recorded at Black Hat 2026, the conversation provides candid insights for founders and security leaders navigating a landscape where building tools is easier than securing organizations.

How it unfolded

• Becherer joined the Early Adopters: CISO podcast, hosted by Front Lines, during the Black Hat 2026 conference.
• He reflected on his career path, including joining Datadog as its sole security employee in 2015.
• The discussion covered the impact of Datadog’s 2016 breach on its long-term security program development.
• Becherer analyzed how agentic coding and AI tools are changing the build-versus-buy decision for security teams.

Who is affected

Security leaders and startup founders are the primary audience for these insights, as they face increasing pressure to secure organizations while managing rapid technological change. According to the podcast summary, Becherer addressed the specific challenges CISOs face when evaluating new technology and deciding which vendors to trust. The conversation also targets professionals responsible for maintaining security tools, highlighting the ongoing responsibilities that come with AI-generated prototypes.

What happens next

The industry continues to grapple with the trade-offs between speed and reliability in security tooling. Becherer emphasized that while AI can accelerate tasks like log analysis and incident-response tool creation, teams must still consider long-term maintenance and changing requirements. For founders, the path forward involves demonstrating practical answers to existing pain points rather than predicting future risks. Becherer shared a specific question he asks vendors to gauge their honesty about security gaps, suggesting that transparency remains key to building confidence in deals.

The background in plain language

Security leaders often struggle to balance rapid innovation with robust protection. When teams build tools using AI, they may create functional prototypes quickly, but those tools require ongoing maintenance, updates, and reliability checks over time. Trust in vendors depends on their ability to address current vulnerabilities honestly rather than promising future fixes. Understanding these dynamics helps leaders evaluate whether a tool will remain secure and supported years after its initial deployment, ensuring long-term organizational safety without relying on unproven predictions.

What to do and how to stay safe: Socket

• Evaluate security vendors based on their transparency about current gaps rather than future promises.
• Assess the long-term maintenance requirements of any AI-built security tools before adoption.
• Ask vendors direct questions about their security posture and how they handle existing vulnerabilities.
• Review whether proposed tools address immediate operational pain points with practical solutions.

Step-by-step guide: Injection attacks examples: how prompt leaks break AI systems

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

Frequently asked questions

What podcast did Andrew Becherer appear on to discuss security leadership?

He appeared on the Early Adopters: CISO podcast, which is part of the Front Lines series.

When and where was this podcast episode recorded?

The episode was recorded at Black Hat 2026, according to the source material.

What company did Becherer join as its first security employee in 2015?

He joined Datadog as its only security employee in 2015, as stated in the summary.