Enterprise AI adoption is accelerating, but moving quickly is only part of the challenge. Organizations must also protect sensitive data, manage access, maintain visibility, and ensure new AI systems can operate securely at scale.

During the AI Networking Summit 2026 in Frisco | Dallas, FedEx Freight leaders Michael Milligan and Prashanth Karne shared how the company is approaching this balance.

Their central message: responsible AI governance should not slow innovation. Done well, it creates a repeatable path for scaling enterprise AI securely and confidently.

What Does It Mean to Scale Enterprise AI Responsibly?

Scaling enterprise AI responsibly means expanding AI use while maintaining clear controls over data, access, risk, accountability, and system performance.

Board-level conversations increasingly focus less on AI as a technology and more on how the organization is managing the risks surrounding it.

As AI becomes embedded across applications and workflows, leaders need to know:

• What data the system can access
• What actions it can take
• Who is responsible for it
• How its activity is monitored
• What controls are in place when something goes wrong

Build AI Governance Around the Business

FedEx Freight created an AI governance committee involving security, legal, finance, and technology leaders. This cross-functional group helps evaluate incoming AI use cases, understand the data involved, and determine whether the right protections are in place.

This type of governance model can help organizations reduce duplicated tools, identify unnecessary risk, and focus resources on AI initiatives that deliver meaningful business value.

The goal is not to create a process that automatically says no. It is to create a safer and more informed path to yes.

Manage AI Agents Like Enterprise Identities

Agentic AI introduces a growing number of non-human identities that may connect to applications, databases, cloud services, and other systems.

Each AI agent should have a defined identity, a clear owner, and only the access required to complete its approved task. Organizations should regularly review those permissions to ensure agents do not become overprivileged.

The familiar principle of least-privilege access is even more important when AI can operate at machine speed.

Observability Is Essential for Responsible AI

Organizations also need visibility into what AI systems are doing after deployment.

Teams should be able to monitor which identities are active, how data is flowing, whether information is moving somewhere unexpected, and whether logs and alerts are sufficient.

FedEx emphasized continuous monitoring as a critical guardrail. Without observability, organizations may not recognize an AI-related issue until end users bring it to their attention.

Speed Still Requires Strong Architecture

AI pilots can be developed quickly, but enterprise deployment requires more than a successful demonstration.

Architecture reviews help determine whether a solution can operate securely, reliably, and sustainably across the organization. Milligan noted that even when leaders want an AI solution immediately, architecture remains essential to long-term security and reliability.

Responsible AI Is an Ongoing Process

AI governance is never finished.

As technologies, use cases, vendors, and risks change, organizations must continue updating their policies, technical controls, monitoring, and employee education.

Karne encouraged technology leaders to keep learning and continually ask what else can be secured, improved, or governed more effectively.

The organizations that scale AI successfully will not be those that eliminate every guardrail. They will be the ones that make responsible innovation repeatable.

Watch the Full FedEx Session On Demand

Watch Moving Fast with AI: How FedEx Scales Innovation with Smart Guardrails to learn how FedEx Freight is approaching enterprise AI governance, identity management, data protection, architecture, and observability.

Watch the full session on demand.

Continue the conversation at the AI Networking Summit NYC, October 28–29, 2026, where enterprise leaders will explore the infrastructure, security, governance, and operating models required to scale AI.

Register for the AI Networking Summit NYC.