The New Security Landscape: AI Agents, Automation, and Human Judgement

The New Security Landscape: AI Agents, Automation, and Human Judgement

From AWS Executive Insights by Amazon Web Services

May 5, 2026 · 21 min

About this episode

The episode discusses the implications of AI agents on security practices and the importance of human judgement in high-stakes environments.

As AI agents become part of day-to-day business operations, security leaders face a new set of questions around identity, autonomy, permissions, and control. In this episode of AWS Executive Insights, Pascal Bornet speaks with Steve Schmidt, Chief Security Officer at Amazon, about how organisations should approach security in an era shaped by AI agents and automation. Bornet and Schmidt discuss why agents should have their own identity, how companies should think about permissions and logging, and why human judgement still matters in high-stakes environments. They also cover the risks introduced by prompt injection, the role of external guardrails, and how AI is already helping defenders move faster against threats at scale. This episode offers a practical view of how security teams and business leaders can respond to the rise of AI agents, with clear actions companies can take now to build safer systems from the start. Learn more about AWS Agentic AI: https://aws.amazon.com/ai/agentic-ai/?trk=80F9DhCw5AKb6zC

People in this episode

Host: Pascal Bornet

Guest: Steve Schmidt

Topics covered

  • AI agents
  • security
  • automation
  • human judgement
  • permissions
  • identity

Keywords

  • AI agents
  • security
  • automation
  • permissions
  • identity
  • prompt injection
  • external guardrails

Mentioned in this episode

Organizations: Amazon, Amazon Web Services

Products: AWS Agentic AI

More episodes of AWS Executive Insights

Explore listener stats, chart rankings, contacts and more on the AWS Executive Insights podcast page.