Leaning Into AI: How Kendal is Approaching Claude Enterprise

The Kendal Corporation is taking a deliberate approach to artificial intelligence. In early June, TKC leadership and Affiliate CEOs gathered for the annual CEO Peer Gathering at LeadingAge in Washington, D.C., to explore how Claude Enterprise can better serve our mission. The conversation began with understanding: 50 years of computing history, where we stand today, and—most importantly—how we use AI in ways that honor and extend who we are.

Starting with values, not features.

The conversation focused on what The Kendal System needs AI to do—and what we don’t need AI to do. When we consider Kendal’s commitment to sustainability, for example, what does it mean to use AI at scale? Leaders learned that Claude Enterprise offers different ways of using the platform—each with different energy demands. Right-sizing the model to the task isn’t just efficient; it’s stewardship. We also discussed using the tool strategically. The goal is to work smarter, not just faster. As we consider how we incorporate AI into our workflow, we must agree on a shared commitment to get the details right and use these tools to improve the excellence of our work.

Understanding the Landscape: Copilot, Claude Enterprise, and ChatGPT

There are multiple AI tools available to users that have different capabilities and use restrictions. As we adopt AI tools, it helps to understand what we’re working with:

APPROVED FOR USE AT KENDAL:

Microsoft Copilot is embedded in Microsoft 365 applications (Word, Excel, Outlook, Teams). Across The Kendal System, a small portion of Kendal System staff are already using Copilot through the Microsoft subscription, and Copilot is built into the apps you use every day.

Claude Enterprise is different. It’s a managed, software-as-a-service (SaaS) platform from Anthropic that connects to Kendal data through a governed pathway of data access controls—not embedded in Office apps. Employees can direct Claude to look at and read files in OneDrive or SharePoint and keep them controlled by permissions identified by Kendal IT, the Kendal System IT group.

Claude Enterprise includes a Business Associate Agreement that allows safe use of protected content (PII, PHI, PFI) within the protected and fully encrypted 365 enclave. This means all of the data is locked at transit and rest. This Anthropic Enterprise Platform is built and ready to be used across the system. It is currently being tested by CEOs while, at the same time, Jonathan Jedeikin, Kendal System Chief Information and Privacy Officer, is working with the System to implement accountable governance and policy around how this technology is used, which includes a process of reporting this governance to each Kendal Affiliate Board of Directors.

NOT APPROVED FOR USE AT KENDAL:

ChatGPT is a consumer tool from OpenAI. It is not supported by The Kendal System’s security policies. Many users may have personal ChatGPT accounts, but this is not recommended within current guidelines.

The key point: Copilot and Claude Enterprise are governed tools with security frameworks and access controls built into their fabrics. ChatGPT on its own does not. If you are seeking to use an AI tool to support your work, you may use Copilot or Claude Enterprise.

State of Play: What this looks like in practice.

CEOs asked a fair question: we’re already busy. How does this actually help?

Morgan Lamphere, Senior Director of Marketing, Sales and Communications, shared her own practice. Every week, she spends 15–20 minutes with Claude Enterprise to prepare for the week ahead. She shares her detailed Outlook calendar, and Claude generates a comprehensive document: an overview of the week, key follow-up areas and a guided to-do list for upcoming meetings and events. One input—her calendar—yields structure, clarity and time back.

Additionally, a new CEO Status Memo was introduced as part of this presentation, which uses Claude and a project management tool to develop a simple, two-page memo for CEOs to guide leadership team meetings and 1:1s.

This is what thoughtful AI adoption looks like: specific to a real need, deployed at a human scale, designed to give time back.

How we’re moving forward: A measured roadmap.

Jonathan Jedeikin (CIPO), has designed an implementation path that is rigorous and grounded:

Training and leadership buy-in (complete): CEOs and leadership teams have been trained. They understand the platform and what it can do.

Governance and policy (in progress): We’re finalizing the privacy and security addendum for CEO consent, followed by collaboration with Kendal System’s HR group.

Organic pilot identification (in progress): CEOs are determining where to start. Some are piloting finance, others culinary operations, facilities or HR. This bottom-up approach means pilots address real operational needs.

Automated workflow development (to come): TKC will develop and implement automation tailored to each pilot.

Observe, feedback, and evolve (to come): We’ll learn from pilots and refine the approach.

This pace is intentional. CEOs are driving adoption at the speed they can sustain. We’re not pushing faster.

Why this matters for the field.

During a recent industry webinar, colleagues from BlueOrange and AQORD shared this: approximately 80% of the senior living industry has no formal AI governance. Kendal does. We’re already implementing everything industry experts recommend. This positions Kendal ahead of the curve—not just in technology adoption, but in the thoughtful, values-aligned way we’re doing it.

The privacy and security framework.

Adoption requires trust. Here’s what the framework protects:

Claude reaches Kendal data only through a governed connector—not via an embedded connection in Office 365.

You can only access files you already have permission to see. Claude inherits your access permissions.

Protected content (PII, PHI, PFI) can be used safely if it stays in OneDrive or SharePoint, where it’s governed. Kendal has a Business Associate Agreement with Anthropic, Claude’s parent company, that covers this use.

Energy-conscious design is built in. The policy guides teams to right-size models and leaves files in place rather than upload them.

Your individual conversations are private. Only the system administrator can export data through a formal process.

Anthropic does not train on Kendal conversations under the Enterprise agreement. This means that Kendal’s prompts, responses and internal information remain separate from the data used to improve Claude’s general models.

A Leadership Perspective.

Adam Day, Chief Executive Officer, Kendal at Lexington, adds:

“Claude Enterprise is already showing us what’s possible. From content creation to data analysis to sales development, teams are finding real ways to work smarter. What matters is that this isn’t a tool imposed from above—Affiliates are leading the way, finding their own uses.”

What’s next.

TKC staff are beginning demonstrations of Claude Enterprise. CEOs will review the privacy and security policy at their next monthly meeting. From there, a phased rollout could unfold with facilitation sessions scheduled as needed. Each team leader will complete pre-work to identify their specific use cases and ensure adherence to the privacy and security framework.

This is how a values-driven organization adopts AI: slowly enough to do it right, deliberately enough to do it well and grounded in the conviction that the tool serves the mission, not the other way around.