June 2026

Whose Risk Is It Anyway? Building the AI Accountability Model

Executive Summary 

Artificial intelligence is rapidly becoming part of everyday healthcare. From clinical documentation and patient engagement to administrative workflows and decision support, AI is helping organizations improve efficiency and accelerate innovation. At the same time, it introduces new questions about accountability, oversight, and risk. 

As AI adoption expands, healthcare organizations are asking a fundamental question: Who owns AI risk? 

In the opening session of Clearwater’s 2026 Healthcare AI Executive Summer Series, moderator Krissy Safi, Senior Vice President of Consulting Services at Clearwater, is joined by Chris Cuellar, Chief Compliance Officer at Elevate ENT, and Blaine Hebert, Vice President and Chief Information Security Officer at Onvida Health, to explore how healthcare organizations are defining AI accountability across the enterprise. 

The discussion makes one point clear: there is no single owner of AI risk. Effective AI governance requires executive leadership, cross-functional collaboration, and clearly defined responsibilities across compliance, cybersecurity, privacy, legal, clinical operations, procurement, and business leadership. Instead of creating entirely new governance models, organizations should build on existing governance and risk management processes while adapting them to address the unique challenges AI introduces. 

Why This Conversation Matters 

Healthcare has reached an important turning point. AI is no longer limited to innovation teams or pilot projects. It is increasingly embedded in electronic health records, clinical documentation tools, patient communication platforms, revenue cycle technologies, cybersecurity solutions, and countless third-party applications. 

This rapid adoption creates tremendous opportunities, but it also raises difficult questions. 

Who evaluates AI vendors? 

Who determines acceptable risk? 

Who monitors AI after deployment? 

Who is accountable when AI influences clinical or operational decisions? 

Throughout the discussion, the panel explains that answering these questions requires organizations to think differently about governance. AI is not simply another technology purchase. It is an enterprise capability that affects nearly every part of a healthcare organization. 

Krissy Safi: “Healthcare has officially crossed a threshold. AI is not sitting in labs anymore.” 

That shift requires governance models that evolve as quickly as the technology itself. 

Key Insights from the Discussion 

AI Risk Is Enterprise Risk 

One of the strongest themes throughout the discussion is that AI risk cannot be assigned to a single department. 

Cybersecurity teams understand technical controls and data protection. Compliance leaders evaluate regulatory obligations. Privacy professionals assess how sensitive information is collected and used. Clinical leaders understand patient safety implications. Legal teams review contractual and liability considerations. Executive leadership ensures AI initiatives align with organizational strategy. 

Each group brings valuable expertise, but none can govern AI independently. 

The panel emphasizes that AI governance succeeds when organizations bring these perspectives together through structured decision-making and shared accountability. 

Shared Governance Requires Clear Ownership 

Although AI governance is collaborative, individual AI initiatives still require defined ownership. 

Chris Cuellar explains that organizations should identify a business owner for every AI implementation. That individual is responsible for driving the initiative while working closely with compliance, cybersecurity, privacy, legal, and operational stakeholders throughout the evaluation and implementation process. 

Chris Cuellar: “It’s a shared risk.” 

Shared governance should never mean unclear accountability. Instead, organizations should establish clear roles while recognizing that AI decisions often affect multiple business functions. 

This approach helps prevent governance gaps while ensuring important decisions are informed by diverse expertise. 

Governance Starts Before AI Is Implemented 

A recurring message throughout the session is that governance should begin well before an AI solution goes live. 

Many organizations wait until implementation to evaluate risk. The panel argues that this is often too late. 

Effective governance begins during procurement by asking questions such as: 

Beginning governance early allows organizations to identify concerns before they become operational challenges. 

It also creates better alignment between technology investments and organizational objectives. 

Executive Leadership Sets the Tone 

AI governance cannot succeed without executive sponsorship. 

Blaine Hebert shares how Onvida Health established a multidisciplinary AI governance committee that includes leaders from across the organization while maintaining direct visibility into executive leadership. 

Blaine Hebert: “We established an AI committee comprised of leaders from across the enterprise.” 

Connecting governance to executive leadership allows organizations to make strategic decisions more effectively while ensuring AI initiatives support broader organizational priorities. 

The discussion reinforces that AI governance is ultimately a business function supported by technology, not the other way around. 

Build on Existing Governance Structures 

One misconception the panel addresses is that organizations need entirely new governance programs for AI. 

Most healthcare organizations already have established processes for enterprise risk management, cybersecurity, vendor management, compliance, privacy, and quality improvement. 

Instead of creating parallel governance structures, AI oversight can often be integrated into these existing processes. 

This approach reduces complexity, improves adoption, and allows organizations to mature governance more quickly. 

The panel encourages organizations to leverage the governance capabilities they already have instead of starting from scratch. 

Responsible AI Requires Continuous Oversight 

Governance does not end after implementation. 

AI systems evolve. Vendors release new capabilities. Organizational priorities change. Regulatory expectations continue to develop. 

The panel explains that organizations should continuously monitor AI performance, reassess risk, review governance processes, and evaluate whether existing controls remain appropriate. 

Successful governance becomes an ongoing operational discipline rather than a one-time approval process. 

Practical Recommendations for Healthcare Leaders 

Throughout the discussion, several practical recommendations emerge for organizations beginning their AI governance journey: 

These recommendations provide a practical foundation for organizations looking to balance innovation with responsible risk management. 

Notable Quotes 

Krissy Safi: “Healthcare has officially crossed a threshold. AI is not sitting in labs anymore.” 

Chris Cuellar: “It’s a shared risk.” 

Blaine Hebert: “We established an AI committee comprised of leaders from across the enterprise.” 

Questions This Discussion Answers 

About This Webinar 

This webinar is the opening session of Clearwater’s 2026 Healthcare AI Executive Summer Series. Krissy Safi moderates a discussion with Chris Cuellar of Elevate ENT and Blaine Hebert of Onvida Health on one of healthcare’s most important AI questions: Who is accountable for AI? Drawing on perspectives from compliance, cybersecurity, and executive leadership, the panel explores how healthcare organizations can establish governance models that support innovation while managing enterprise risk. Whether your organization is beginning to adopt AI or expanding existing governance efforts, this discussion provides practical guidance for building accountability, improving cross-functional collaboration, and creating governance programs that can evolve alongside the future of healthcare AI.