Why They Must Be Separate but Coordinated
The IT department and the AI department serve different purposes. The separation is not about hierarchy. It is about accountability.
IT exists to keep the organization stable. Its focus is hardware, networking, identity, permissions, security controls, patching, vendor software maintenance, and uptime. IT standardizes systems so the company can operate predictably. Much of IT work is driven by user error or misuse of static tools. These tools do not make decisions. They behave the same way each time.
AI systems are different.
AI systems generate outputs. They make recommendations. Their behavior can change over time. A model can update without a visible software update. A vendor can enable new features by default. Outputs can drift. Results can change even if no one in the company deployed anything new.
These are not uptime problems. These are behavioral and governance problems.
What the AI Department Owns
The AI department owns evaluation, configuration oversight, and behavioral governance.
- Determines whether a system should be used
- Defines how models are configured or trained
- Decides what data can be included and what must be excluded
- Monitors output quality and drift
- Works with vendors when behavior changes
- Defines ethical and regulatory guardrails
- Teaches departments how to use systems correctly
It does not control hardware or network infrastructure.
What IT Owns
IT owns the technical environment.
- Defines approved infrastructure and technical constraints
- Controls identity, access, and permissions
- Enforces security controls and operational standards
- Manages logging, monitoring, and uptime
- Handles deployment and vendor software maintenance
- Supports user access issues
If a system cannot operate within the environment, it does not move forward.
Where They Interact
- Before rollout: The AI department evaluates behavior, risk, and compliance. IT validates that the system can operate securely inside the company environment.
- During operation: IT maintains production stability. The AI department monitors outputs and behavioral impact.
- When behavior changes: If a vendor updates a model or changes features, the AI department reviews behavioral impact. IT ensures access, containment, and rollback capability.
Incident Responsibility
- AI department: If an AI system produces harmful or non compliant output, the AI department owns the behavioral correction. This includes configuration changes, data adjustments, policy updates, and vendor escalation.
- IT: IT owns containment. IT restricts access if needed. IT manages rollback. IT preserves production stability.
- Joint: Post incident review is joint. Accountability is defined.
AI Operations Specialists
AI Operations Specialists operate within business units and coordinate directly with the AI department.
- Understand departmental workflows and ensure AI systems are used according to defined standards
- Monitor output quality within their function and surface behavioral concerns early
- Coordinate with IT on deployment timing, access alignment, and workflow integration
- Escalate technical issues to IT
- Escalate behavioral or output issues to the AI department
They do not manage infrastructure, access control, user rights, or core security enforcement. Those remain IT responsibilities. However, they coordinate closely with IT on AI related security configurations. If a model feature should be disabled, logging increased, or data boundaries adjusted, those decisions are made jointly. IT enforces the controls within the environment. The AI department defines the behavioral and risk requirements driving those controls.
Summary
- IT owns systems.
- The AI department owns behavioral governance and oversight.
- Production responsibility is shared with clear boundaries.
This structure creates clarity. It avoids burying AI governance inside a support function. It also prevents AI from operating outside enterprise controls.