Agentic AI in Health-Tech and Healthcare, Part 2

Drawing from Straiker deployment, assessment, and technical evaluation records across dozens of healthcare and health-tech organizations, this report looks at what security teams are actually encountering as they secure AI agents, coding tools, and increasingly complex agentic environments.

Part 1 looked at where AI is showing up across healthcare. Part 2 goes inside the security program. Inside the report:

• See which AI systems healthcare security teams are prioritizing first, including coding agents and employee copilots.
• Understand why healthcare agent stacks are harder to inventory and govern across multiple platforms, tools, and teams.
• Explore the gap between discovery and enforcement, and why visibility alone is not enough.
• Learn why realistic adversarial testing is harder in healthcare, especially when production systems carry real data and consequences.
• See what strong runtime enforcement and containment should look like when an agent can no longer be trusted.
• Use a practical buyer’s checklist for evaluating agentic AI security controls in healthcare.

Part 1 showed where AI is being used. Part 2 shows what security teams are finding once they start trying to secure it.