Check AI Governance as Healthcare Cyberattacks Disrupt Patient Care

Healthcare system administrators must review their AI governance controls immediately, as a new report links unchecked AI tools to significant data loss and patient care disruptions. According to research published by Proofpoint and the Ponemon Institute, cyberattacks are no longer just IT incidents; they are directly impacting clinical outcomes and operational stability across the United States. The findings, released in October 2026, show that nearly all surveyed organizations faced attacks, with a majority experiencing tangible harm to patient services.

In plain English

The primary issue is that cyberattacks are causing direct harm to patients, not just financial losses. The report, based on 705 IT and security practitioners, found that 97% of organizations experienced at least one attack in the last 12 months. Of those, 74% reported that these attacks disrupted patient care. This disruption manifests as delayed medical procedures, extended hospital stays, and increased complications. The average number of attacks per organization also climbed to 51, up from 43 in 2025, indicating a worsening trend in frequency and intensity.

The background

Financial costs are rising alongside operational impact. The average cost of the most expensive cyberattack increased by nearly 8% to $4.2 million in 2026. Disruption to normal healthcare operations accounted for the largest single cost category at $1.39 million, followed by $1.05 million from idle time and lost productivity. Supply chain attacks proved particularly damaging, with 90% of organizations experiencing these attacks reporting care disruption. Ransomware was most associated with longer patient stays, while business email compromise was linked to delays in tests and procedures.

What changes now

AI adoption is outpacing security governance, creating new vulnerabilities. While 45% of organizations have embedded AI in cybersecurity or patient care, only 42% use security measures specifically designed for AI systems. Sixty percent lack controls governing sensitive data used in AI systems, and just 32% have full visibility into employees’ use of unsanctioned AI tools. Consequently, 24% of respondents cited AI agents operating without guardrails as a leading cause of data loss. Supply chain risk from AI vendors, data leakage through prompts, and AI-generated phishing are the top AI-specific concerns.

What to do and how to stay safe: Proofpoint

• Audit all AI tools used by staff to ensure they have appropriate guardrails and do not process sensitive patient data without authorization.
• Review supply chain security for third-party vendors, especially those providing AI services, as 40% of organizations cite this as a leading concern.
• Implement visibility controls to monitor employee use of unsanctioned AI tools, since only 32% of organizations currently have full oversight.
• Assess incident response plans to specifically address delays in medical procedures and tests, which are common consequences of ransomware and BEC attacks.

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

Frequently asked questions

How much did the average cost of cyberattacks in healthcare rise?

The average cost of the most expensive cyberattack rose nearly 8% to $4.2 million in 2026, up from $3.9 million in 2025.

What percentage of healthcare organizations experienced patient care disruption?

According to the Proofpoint-Ponemon Institute report, 74% of surveyed US healthcare organizations reported disruption to patient care due to cyberattacks.

What are the main AI-related security concerns for healthcare?

Leading concerns include supply chain risk from AI vendors (40%), data leakage through prompts (36%), and AI-generated phishing and social engineering (34%).