DNSFilter Survey Finds 43% of Healthcare Security Leaders Name Employees Sharing Sensitive Data With Public AI as Their Biggest Emerging Risk
DNSFilter Survey Finds 43% of Healthcare Security Leaders Name Employees Sharing Sensitive Data With Public AI as Their Biggest Emerging Risk
- 60% of healthcare organizations reply on vendor security notifications to learn of AI and SaaS incidents, while only 23% identify them through employee reporting
- Half of healthcare organizations have confirmed or suspect an unapproved AI connection is being made to critical systems, yet only 40% consistently enforce their own internal AI policies
- 87% of healthcare security leaders are confident they would detect a breach within six hours, yet one in three leaders has had an incident that was never formally reported or disclosed
WASHINGTON--(BUSINESS WIRE)--DNSFilter, the DNS-first security company, powered by AI, today released Seeing the Risk, Missing the Signal: AI Security in Healthcare, the latest report in its 2026 Visibility Deficit cybersecurity survey series, revealing that 43% of healthcare security leaders say employees sharing sensitive data with public AI models is the biggest emerging AI risk over the next 12 months. Only 23% of healthcare organizations identify SaaS and AI incidents through employee reporting, while 60% rely on vendor security notifications to be informed. The study found that this visibility gap runs through how incidents are detected, whether AI policies are enforced, and where security budget is going.
Healthcare is clear about where AI creates exposure. More than half of respondents (53%) name the loss of intellectual property through AI training data as a top AI risk, while half call AI security and governance one of their biggest security gaps. However, visibility is where that awareness falls short. Only 27% can say every security tool they own is actively monitored, and 40% say half or fewer are.
That blind spot is already letting unapproved AI through. Half of healthcare organizations have either confirmed (27%) or suspect (23%) that an employee has connected an unapproved AI tool to Google Workspace, Microsoft 365, or their SSO environment. Written policy is not stopping it either: 70% have a formal AI policy, but only 40% enforce it consistently.
"Healthcare security leaders know exactly where AI puts their data at risk. What they’re missing is the moment that risk becomes an incident, and too many are hearing about it from a vendor instead of their own team. When the data involved is protected health information, the risk only becomes more significant. Before the next dollar goes to AI governance, healthcare organizations need to see which AI services their people are already using without clearance," said Ken Carnesi, CEO, DNSFilter.
Confidence in healthcare remains strong, even while incident reporting continues to miss the mark. While 87% of healthcare security leaders are confident they would detect a breach within six hours, one in three had an incident originate from a third-party SaaS or AI tool in the past year, and the same share have had an incident that was never formally reported or disclosed.
Spending has not caught up either. Healthcare names AI governance as its biggest security gap, yet only 33% say their next security dollar will go to AI security and governance. The stakes are high: 57% could not maintain critical operations through a 72-hour disruption, and in healthcare, an outage of that length can not only delay patient care but risk lives.
“Most AI risk in a healthcare organization doesn’t look like someone opening a chatbot. It looks like a browser extension, a plug-in, or an OAuth grant approved in a few clicks, and none of that shows up in a policy document,” said Justin Foster, CTO, DNSFilter. “What those tools have in common is that they all have to look up a domain before they can move any data. Watching that layer shows healthcare security teams what’s actually connecting to their systems and gives them a way to enforce the AI policy they’ve already written.”
Additional key findings from the report include:
- 47% of healthcare organizations allow employees to grant a new AI tool OAuth or API access on their own
- Only 13% name AI-enabled phishing as a top emerging risk
The DNSFilter AI Cybersecurity Survey 2026 was conducted among 400 IT and cybersecurity professionals at the manager level and above, at companies with 150 to 5,000 employees, across North America, in June 2026. Read the full report and findings here.
More than 57,000 customers worldwide rely on DNSFilter, from global enterprises and mid-market companies to educational institutions, government agencies, MSPs and MSSPs. DNSFilter reaches them directly and through a global partner ecosystem that includes CrowdStrike, Amazon, GuidePoint Security, Climb, SHI, Pax8, N-able and Midis Group.
About DNSFilter
DNSFilter, the DNS-first security company, powered by AI, protects every user at the first click, preventing threats before a connection is ever made, securing organizations and their people wherever they work, live and travel.
DNSFilter stops cyber criminals, rogue states and other malicious actors from hacking your world by delivering best-in-class DNS filtering; behavior analytics; AI governance and controls; encryption and privacy and threat intelligence services that are built upon its proprietary DNS-first platform and intelligence engine.
Its LUMINATE by DNSFilter™ platform is trusted by more than 57,000 organizations around the world, helping to protect thousands of global enterprises, mid-market companies, educational institutions, agencies and MSPs that operate across government, transportation, retail, hospitality, healthcare and many more verticals. 100 million+ users worldwide rely on DNSFilter to keep them safe every day. Deployed in minutes, it protects users at the first source of contact. Learn more at dnsfilter.com.
Contacts
Media Contact:
Rebecca Cradick
Vice President, Global Communications and Investor Relations
media@dnsfilter.com
